Index: head/en_US.ISO8859-1/books/porters-handbook/makefiles/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/porters-handbook/makefiles/chapter.xml (revision 48371) +++ head/en_US.ISO8859-1/books/porters-handbook/makefiles/chapter.xml (revision 48372) @@ -1,5429 +1,5445 @@ Configuring the Makefile Configuring the Makefile is pretty simple, and again we suggest looking at existing examples before starting. Also, there is a sample Makefile in this handbook, so take a look and please follow the ordering of variables and sections in that template to make the port easier for others to read. Consider these problems in sequence during the design of the new Makefile: The Original Source Does it live in DISTDIR as a standard gzipped tarball named something like foozolix-1.2.tar.gz? If so, go on to the next step. If not, the distribution file format might require overriding one or more of DISTVERSION, DISTNAME, EXTRACT_CMD, EXTRACT_BEFORE_ARGS, EXTRACT_AFTER_ARGS, EXTRACT_SUFX, or DISTFILES. In the worst case, create a custom do-extract target to override the default. This is rarely, if ever, necessary. Naming The first part of the port's Makefile names the port, describes its version number, and lists it in the correct category. <varname>PORTNAME</varname> and <varname>PORTVERSION</varname> Set PORTNAME to the base name of the port. Set PORTVERSION to the version number of the port unless DISTVERSION is used (see ). The package name must be unique across the entire ports tree. Make sure that the PORTNAME is not already in use by an existing port. If the name has already been used, add either PKGNAMEPREFIX or PKGNAMESUFFIX. <varname>PORTREVISION</varname> and <varname>PORTEPOCH</varname> <varname>PORTREVISION</varname> PORTREVISION is a monotonically increasing value which is reset to 0 with every increase of PORTVERSION, typically every time there is a new official vendor release. If PORTREVISION is non-zero, the value is appended to the package name. Changes to PORTREVISION are used by automated tools like &man.pkg-version.8; to determine that a new package is available. PORTREVISION must be increased each time a change is made to the port that changes the generated package in any way. That includes changes that only affect a package built with non-default options. Examples of when PORTREVISION must be bumped: Addition of patches to correct security vulnerabilities, bugs, or to add new functionality to the port. Changes to the port Makefile to enable or disable compile-time options in the package. Changes in the packing list or the install-time behavior of the package. For example, a change to a script which generates initial data for the package, like &man.ssh.1; host keys. Version bump of a port's shared library dependency (in this case, someone trying to install the old package after installing a newer version of the dependency will fail since it will look for the old libfoo.x instead of libfoo.(x+1)). Silent changes to the port distfile which have significant functional differences. For example, changes to the distfile requiring a correction to distinfo with no corresponding change to PORTVERSION, where a diff -ru of the old and new versions shows non-trivial changes to the code. Examples of changes which do not require a PORTREVISION bump: Style changes to the port skeleton with no functional change to what appears in the resulting package. Changes to MASTER_SITES or other functional changes to the port which do not affect the resulting package. Trivial patches to the distfile such as correction of typos, which are not important enough that users of the package have to go to the trouble of upgrading. Build fixes which cause a package to become compilable where it was previously failing. As long as the changes do not introduce any functional change on any other platforms on which the port did previously build. Since PORTREVISION reflects the content of the package, if the package was not previously buildable then there is no need to increase PORTREVISION to mark a change. A rule of thumb is to decide whether a change committed to a port is something which some people would benefit from having. Either because of an enhancement, fix, or by virtue that the new package will actually work at all. Then weigh that against that fact that it will cause everyone who regularly updates their ports tree to be compelled to update. If yes, PORTREVISION must be bumped. People using binary packages will never see the update if PORTREVISION is not bumped. Without increasing PORTREVISION, the package builders have no way to detect the change and thus, will not rebuild the package. <varname>PORTEPOCH</varname> From time to time a software vendor or &os; porter will do something silly and release a version of their software which is actually numerically less than the previous version. An example of this is a port which goes from foo-20000801 to foo-1.0 (the former will be incorrectly treated as a newer version since 20000801 is a numerically greater value than 1). The results of version number comparisons are not always obvious. pkg version (see &man.pkg-version.8;) can be used to test the comparison of two version number strings. For example: &prompt.user; pkg version -t 0.031 0.29 > The > output indicates that version 0.031 is considered greater than version 0.29, which may not have been obvious to the porter. In situations such as this, PORTEPOCH must be increased. If PORTEPOCH is nonzero it is appended to the package name as described in section 0 above. PORTEPOCH must never be decreased or reset to zero, because that would cause comparison to a package from an earlier epoch to fail. For example, the package would not be detected as out of date. The new version number, 1.0,1 in the above example, is still numerically less than the previous version, 20000801, but the ,1 suffix is treated specially by automated tools and found to be greater than the implied suffix ,0 on the earlier package. Dropping or resetting PORTEPOCH incorrectly leads to no end of grief. If the discussion above was not clear enough, please consult the &a.ports;. It is expected that PORTEPOCH will not be used for the majority of ports, and that sensible use of PORTVERSION can often preempt it becoming necessary if a future release of the software changes the version structure. However, care is needed by &os; porters when a vendor release is made without an official version number — such as a code snapshot release. The temptation is to label the release with the release date, which will cause problems as in the example above when a new official release is made. For example, if a snapshot release is made on the date 20000917, and the previous version of the software was version 1.2, do not use 20000917 for PORTVERSION. The correct way is a PORTVERSION of 1.2.20000917, or similar, so that the succeeding release, say 1.3, is still a numerically greater value. Example of <varname>PORTREVISION</varname> and <varname>PORTEPOCH</varname> Usage The gtkmumble port, version 0.10, is committed to the ports collection: PORTNAME= gtkmumble PORTVERSION= 0.10 PKGNAME becomes gtkmumble-0.10. A security hole is discovered which requires a local &os; patch. PORTREVISION is bumped accordingly. PORTNAME= gtkmumble PORTVERSION= 0.10 PORTREVISION= 1 PKGNAME becomes gtkmumble-0.10_1 A new version is released by the vendor, numbered 0.2 (it turns out the author actually intended 0.10 to actually mean 0.1.0, not what comes after 0.9 - oops, too late now). Since the new minor version 2 is numerically less than the previous version 10, PORTEPOCH must be bumped to manually force the new package to be detected as newer. Since it is a new vendor release of the code, PORTREVISION is reset to 0 (or removed from the Makefile). PORTNAME= gtkmumble PORTVERSION= 0.2 PORTEPOCH= 1 PKGNAME becomes gtkmumble-0.2,1 The next release is 0.3. Since PORTEPOCH never decreases, the version variables are now: PORTNAME= gtkmumble PORTVERSION= 0.3 PORTEPOCH= 1 PKGNAME becomes gtkmumble-0.3,1 If PORTEPOCH were reset to 0 with this upgrade, someone who had installed the gtkmumble-0.10_1 package would not detect the gtkmumble-0.3 package as newer, since 3 is still numerically less than 10. Remember, this is the whole point of PORTEPOCH in the first place. <varname>PKGNAMEPREFIX</varname> and <varname>PKGNAMESUFFIX</varname> Two optional variables, PKGNAMEPREFIX and PKGNAMESUFFIX, are combined with PORTNAME and PORTVERSION to form PKGNAME as ${PKGNAMEPREFIX}${PORTNAME}${PKGNAMESUFFIX}-${PORTVERSION}. Make sure this conforms to our guidelines for a good package name. In particular, the use of a hyphen (-) in PORTVERSION is not allowed. Also, if the package name has the language- or the -compiled.specifics part (see below), use PKGNAMEPREFIX and PKGNAMESUFFIX, respectively. Do not make them part of PORTNAME. Package Naming Conventions These are the conventions to follow when naming packages. This is to make the package directory easy to scan, as there are already thousands of packages and users are going to turn away if they hurt their eyes! Package names take the form of language_region-name-compiled.specifics-version.numbers. The package name is defined as ${PKGNAMEPREFIX}${PORTNAME}${PKGNAMESUFFIX}-${PORTVERSION}. Make sure to set the variables to conform to that format. language_region- &os; strives to support the native language of its users. The language- part is a two letter abbreviation of the natural language defined by ISO-639 when the port is specific to a certain language. Examples are ja for Japanese, ru for Russian, vi for Vietnamese, zh for Chinese, ko for Korean and de for German. If the port is specific to a certain region within the language area, add the two letter country code as well. Examples are en_US for US English and fr_CH for Swiss French. The language- part is set in PKGNAMEPREFIX. name Make sure that the port's name and version are clearly separated and placed into PORTNAME and PORTVERSION. The only reason for PORTNAME to contain a version part is if the upstream distribution is really named that way, as in the textproc/libxml2 or japanese/kinput2-freewnn ports. Otherwise, PORTNAME cannot contain any version-specific information. It is quite normal for several ports to have the same PORTNAME, as the www/apache* ports do; in that case, different versions (and different index entries) are distinguished by PKGNAMEPREFIX and PKGNAMESUFFIX values. There is a tradition of naming Perl 5 modules by prepending p5- and converting the double-colon separator to a hyphen. For example, the Data::Dumper module becomes p5-Data-Dumper. -compiled.specifics If the port can be built with different hardcoded defaults (usually part of the directory name in a family of ports), the -compiled.specifics part states the compiled-in defaults. The hyphen is optional. Examples are paper size and font units. The -compiled.specifics part is set in PKGNAMESUFFIX. -version.numbers The version string follows a dash (-) and is a period-separated list of integers and single lowercase alphabetics. In particular, it is not permissible to have another dash inside the version string. The only exception is the string pl (meaning patchlevel), which can be used only when there are no major and minor version numbers in the software. If the software version has strings like alpha, beta, rc, or pre, take the first letter and put it immediately after a period. If the version string continues after those names, the numbers follow the single alphabet without an extra period between them (for example, 1.0b2). The idea is to make it easier to sort ports by looking at the version string. In particular, make sure version number components are always delimited by a period, and if the date is part of the string, use the 0.0.yyyy.mm.dd format, not dd.mm.yyyy or the non-Y2K compliant yy.mm.dd format. It is important to prefix the version with 0.0. in case a release with an actual version number is made, which would be numerically less than yyyy. Package name must be unique among all of the ports tree, check that there is not already a port with the same PORTNAME and if there is add one of PKGNAMEPREFIX or PKGNAMESUFFIX. Here are some (real) examples on how to convert the name as called by the software authors to a suitable package name: Package Naming Examples Distribution Name PKGNAMEPREFIX PORTNAME PKGNAMESUFFIX PORTVERSION Reason mule-2.2.2 (empty) mule (empty) 2.2.2 No changes required mule-1.0.1 (empty) mule 1 1.0.1 mule already exists EmiClock-1.0.2 (empty) emiclock (empty) 1.0.2 No uppercase names for single programs rdist-1.3alpha (empty) rdist (empty) 1.3.a No strings like alpha allowed es-0.9-beta1 (empty) es (empty) 0.9.b1 No strings like beta allowed mailman-2.0rc3 (empty) mailman (empty) 2.0.r3 No strings like rc allowed v3.3beta021.src (empty) tiff (empty) 3.3 What the heck was that anyway? tvtwm (empty) tvtwm (empty) pl11 Version string always required piewm (empty) piewm (empty) 1.0 Version string always required xvgr-2.10pl1 (empty) xvgr (empty) 2.10.1 pl allowed only when no major/minor version numbers gawk-2.15.6 ja- gawk (empty) 2.15.6 Japanese language version psutils-1.13 (empty) psutils -letter 1.13 Paper size hardcoded at package build time pkfonts (empty) pkfonts 300 1.0 Package for 300dpi fonts
If there is absolutely no trace of version information in the original source and it is unlikely that the original author will ever release another version, just set the version string to 1.0 (like the piewm example above). Otherwise, ask the original author or use the date string the source file was released on (0.0.yyyy.mm.dd) as the version. Categorization <varname>CATEGORIES</varname> When a package is created, it is put under /usr/ports/packages/All and links are made from one or more subdirectories of /usr/ports/packages. The names of these subdirectories are specified by the variable CATEGORIES. It is intended to make life easier for the user when he is wading through the pile of packages on the FTP site or the CDROM. Please take a look at the current list of categories and pick the ones that are suitable for the port. This list also determines where in the ports tree the port is imported. If there is more than one category here, the port files must be put in the subdirectory with the name of the first category. See below for more discussion about how to pick the right categories. Current List of Categories Here is the current list of port categories. Those marked with an asterisk (*) are virtual categories—those that do not have a corresponding subdirectory in the ports tree. They are only used as secondary categories, and only for search purposes. For non-virtual categories, there is a one-line description in COMMENT in that subdirectory's Makefile. Category Description Notes accessibility Ports to help disabled users. afterstep* Ports to support the AfterStep window manager. arabic Arabic language support. archivers Archiving tools. astro Astronomical ports. audio Sound support. benchmarks Benchmarking utilities. biology Biology-related software. cad Computer aided design tools. chinese Chinese language support. comms Communication software. Mostly software to talk to the serial port. converters Character code converters. databases Databases. deskutils Things that used to be on the desktop before computers were invented. devel Development utilities. Do not put libraries here just because they are libraries. They should not be in this category unless they truly do not belong anywhere else. dns DNS-related software. docs* Meta-ports for &os; documentation. editors General editors. Specialized editors go in the section for those tools. For example, a mathematical-formula editor will go in math, and have editors as a second category. elisp* Emacs-lisp ports. emulators Emulators for other operating systems. Terminal emulators do not belong here. X-based ones go to x11 and text-based ones to either comms or misc, depending on the exact functionality. finance Monetary, financial and related applications. french French language support. ftp FTP client and server utilities. If the port speaks both FTP and HTTP, put it in ftp with a secondary category of www. games Games. geography* Geography-related software. german German language support. gnome* Ports from the GNOME Project. gnustep* Software related to the GNUstep desktop environment. graphics Graphics utilities. hamradio* Software for amateur radio. haskell* Software related to the Haskell language. hebrew Hebrew language support. hungarian Hungarian language support. ipv6* IPv6 related software. irc Internet Relay Chat utilities. japanese Japanese language support. java Software related to the Java™ language. The java category must not be the only one for a port. Save for ports directly related to the Java language, porters are also encouraged not to use java as the main category of a port. kde* Ports from the KDE Project. kld* Kernel loadable modules. korean Korean language support. lang Programming languages. linux* Linux applications and support utilities. lisp* Software related to the Lisp language. mail Mail software. math Numerical computation software and other utilities for mathematics. mbone* MBone applications. misc Miscellaneous utilities Things that do not belong anywhere else. If at all possible, try to find a better category for the port than misc, as ports tend to be overlooked in here. multimedia Multimedia software. net Miscellaneous networking software. net-im Instant messaging software. net-mgmt Networking management software. net-p2p Peer to peer network applications. news USENET news software. palm Software support for the Palm™ series. parallel* Applications dealing with parallelism in computing. pear* Ports related to the Pear PHP framework. perl5* Ports that require Perl version 5 to run. plan9* Various programs from Plan9. polish Polish language support. ports-mgmt Ports for managing, installing and developing &os; ports and packages. portuguese Portuguese language support. print Printing software. Desktop publishing tools (previewers, etc.) belong here too. python* Software related to the Python language. ruby* Software related to the Ruby language. rubygems* Ports of RubyGems packages. russian Russian language support. scheme* Software related to the Scheme language. science Scientific ports that do not fit into other categories such as astro, biology and math. security Security utilities. shells Command line shells. spanish* Spanish language support. sysutils System utilities. tcl* Ports that use Tcl to run. textproc Text processing utilities. It does not include desktop publishing tools, which go to print. tk* Ports that use Tk to run. ukrainian Ukrainian language support. vietnamese Vietnamese language support. windowmaker* Ports to support the WindowMaker window manager. www Software related to the World Wide Web. HTML language support belongs here too. x11 The X Window System and friends. This category is only for software that directly supports the window system. Do not put regular X applications here. Most of them go into other x11-* categories (see below). x11-clocks X11 clocks. x11-drivers X11 drivers. x11-fm X11 file managers. x11-fonts X11 fonts and font utilities. x11-servers X11 servers. x11-themes X11 themes. x11-toolkits X11 toolkits. x11-wm X11 window managers. xfce* Ports related to the Xfce desktop environment. zope* Zope support. Choosing the Right Category As many of the categories overlap, choosing which of the categories will be the primary category of the port can be tedious. There are several rules that govern this issue. Here is the list of priorities, in decreasing order of precedence: The first category must be a physical category (see above). This is necessary to make the packaging work. Virtual categories and physical categories may be intermixed after that. Language specific categories always come first. For example, if the port installs Japanese X11 fonts, then the CATEGORIES line would read japanese x11-fonts. Specific categories are listed before less-specific ones. For instance, an HTML editor is listed as www editors, not the other way around. Also, do not list net when the port belongs to any of irc, mail, news, security, or www, as net is included implicitly. x11 is used as a secondary category only when the primary category is a natural language. In particular, do not put x11 in the category line for X applications. Emacs modes are placed in the same ports category as the application supported by the mode, not in editors. For example, an Emacs mode to edit source files of some programming language goes into lang. Ports installing loadable kernel modules also have the virtual category kld in their CATEGORIES line. This is one of the things handled automatically by adding USES=kmod. misc does not appear with any other non-virtual category. If there is misc with something else in CATEGORIES, that means misc can safely be deleted and the port placed only in the other subdirectory. If the port truly does not belong anywhere else, put it in misc. If the category is not clearly defined, please put a comment to that effect in the port submission in the bug database so we can discuss it before we import it. As a committer, send a note to the &a.ports; so we can discuss it first. Too often, new ports are imported to the wrong category only to be moved right away. This causes unnecessary and undesirable bloat in the master source repository. Proposing a New Category As the Ports Collection has grown over time, various new categories have been introduced. New categories can either be virtual categories—those that do not have a corresponding subdirectory in the ports tree— or physical categories—those that do. This section discusses the issues involved in creating a new physical category. Read it thouroughly before proposing a new one. Our existing practice has been to avoid creating a new physical category unless either a large number of ports would logically belong to it, or the ports that would belong to it are a logically distinct group that is of limited general interest (for instance, categories related to spoken human languages), or preferably both. The rationale for this is that such a change creates a fair amount of work for both the committers and also for all users who track changes to the Ports Collection. In addition, proposed category changes just naturally seem to attract controversy. (Perhaps this is because there is no clear consensus on when a category is too big, nor whether categories should lend themselves to browsing (and thus what number of categories would be an ideal number), and so forth.) Here is the procedure: Propose the new category on &a.ports;. Include a detailed rationale for the new category, including why the existing categories are not sufficient, and the list of existing ports proposed to move. (If there are new ports pending in Bugzilla that would fit this category, list them too.) If you are the maintainer and/or submitter, respectively, mention that as it may help the case. Participate in the discussion. If it seems that there is support for the idea, file a PR which includes both the rationale and the list of existing ports that need to be moved. Ideally, this PR would also include these patches: Makefiles for the new ports once they are repocopied Makefile for the new category Makefile for the old ports' categories Makefiles for ports that depend on the old ports (for extra credit, include the other files that have to change, as per the procedure in the Committer's Guide.) Since it affects the ports infrastructure and involves moving and patching many ports but also possibly running regression tests on the build cluster, assign the PR to the &a.portmgr;. If that PR is approved, a committer will need to follow the rest of the procedure that is outlined in the Committer's Guide. Proposing a new virtual category is similar to the above but much less involved, since no ports will actually have to move. In this case, the only patches to include in the PR would be those to add the new category to CATEGORIES of the affected ports. Proposing Reorganizing All the Categories Occasionally someone proposes reorganizing the categories with either a 2-level structure, or some other kind of keyword structure. To date, nothing has come of any of these proposals because, while they are very easy to make, the effort involved to retrofit the entire existing ports collection with any kind of reorganization is daunting to say the very least. Please read the history of these proposals in the mailing list archives before posting this idea. Furthermore, be prepared to be challenged to offer a working prototype. The Distribution Files The second part of the Makefile describes the files that must be downloaded to build the port, and where they can be downloaded. <varname>DISTVERSION/DISTNAME</varname> DISTNAME is the name of the port as called by the authors of the software. DISTNAME defaults to ${PORTNAME}-${DISTVERSIONPREFIX}${DISTVERSION}${DISTVERSIONSUFFIX}, and DISTVERSION defaults to ${PORTVERSION} so override it only if necessary. DISTNAME is only used in two places. First, the distribution file list (DISTFILES) defaults to ${DISTNAME}${EXTRACT_SUFX}. Second, the distribution file is expected to extract into a subdirectory named WRKSRC, which defaults to work/${DISTNAME}. Some vendor's distribution names which do not fit into the ${PORTNAME}-${PORTVERSION}-scheme can be handled automatically by setting DISTVERSION. PORTVERSION will be derived from it automatically. Only one of PORTVERSION and DISTVERSION can be set at a time. If DISTVERSION does not derive a correct PORTVERSION, do not use DISTVERSION, set PORTVERSION to the right value and set DISTNAME with PORTNAME with either some computation of PORTVERSION or the verbatim upstream version. Examples of <varname>DISTVERSION</varname> and the Derived <varname>PORTVERSION</varname> DISTVERSION PORTVERSION 0.7.1d 0.7.1.d 10Alpha3 10.a3 3Beta7-pre2 3.b7.p2 8:f_17 8f.17
PKGNAMEPREFIX and PKGNAMESUFFIX do not affect DISTNAME. Also note that if WRKSRC is equal to ${WRKDIR}/${DISTNAME} while the original source archive is named something other than ${PORTNAME}-${PORTVERSION}${EXTRACT_SUFX}, leave DISTNAME alone— defining only DISTFILES is easier than both DISTNAME and WRKSRC (and possibly EXTRACT_SUFX). <varname>MASTER_SITES</varname> Record the directory part of the FTP/HTTP-URL pointing at the original tarball in MASTER_SITES. Do not forget the trailing slash (/)! The make macros will try to use this specification for grabbing the distribution file with FETCH if they cannot find it already on the system. It is recommended that multiple sites are included on this list, preferably from different continents. This will safeguard against wide-area network problems. We are even planning to add support for automatically determining the closest master site and fetching from there; having multiple sites will go a long way towards helping this effort. MASTER_SITES must not be blank. It must point to the actual site hosting the distribution files. It cannot point to web archives, or the &os; distribution files cache sites. The only exception to this rule is ports that do not have any distribution files. For example, meta-ports do not have any distribution files, so MASTER_SITES does not need to be set. Using <varname>MASTER_SITE_<replaceable>*</replaceable></varname> Variables Shortcut abbreviations are available for popular archives like SourceForge (SOURCEFORGE), GNU (GNU), or Perl CPAN (PERL_CPAN). MASTER_SITES can use them directly: MASTER_SITES= GNU/make The older expanded format still works, but all ports have been converted to the compact format. The expanded format looks like this: MASTER_SITES= ${MASTER_SITE_GNU} MASTER_SITE_SUBDIR= make These values and variables are defined in Mk/bsd.sites.mk. New entries are added often, so make sure to check the latest version of this file before submitting a port. For any MASTER_SITE_FOO variable, the shorthand FOO can be used. For example, use: MASTER_SITES= FOO If MASTER_SITE_SUBDIR is needed, use this: MASTER_SITES= FOO/bar Some MASTER_SITE_* names are quite long, and for ease of use, shortcuts have been defined: Shortcuts for <varname>MASTER_SITE_<replaceable>*</replaceable></varname> Macros Macro Shortcut PERL_CPAN CPAN GITHUB GH GITHUB_CLOUD GHC LIBREOFFICE_DEV LODEV NETLIB NL RUBYGEMS RG SOURCEFORGE SF SOURCEFORGE_JP SFJP
Magic MASTER_SITES Macros Several magic macros exist for popular sites with a predictable directory structure. For these, just use the abbreviation and the system will choose a subdirectory automatically. For a port named Stardict, of version 1.2.3, and hosted on SourceForge, adding this line: MASTER_SITES= SF infers a subdirectory named /project/stardict/stardict/1.2.3. If the inferred directory is incorrect, it can be overridden: MASTER_SITES= SF/stardict/WyabdcRealPeopleTTS/${PORTVERSION} This can also be written as MASTER_SITES= SF MASTER_SITE_SUBDIR= stardict/WyabdcRealPeopleTTS/${PORTVERSION} Magic <varname>MASTER_SITES</varname> Macros Macro Assumed subdirectory APACHE_COMMONS_BINARIES ${PORTNAME:S,commons-,,} APACHE_COMMONS_SOURCE ${PORTNAME:S,commons-,,} APACHE_JAKARTA ${PORTNAME:S,-,/,}/source BERLIOS ${PORTNAME:tl}.berlios CHEESESHOP source/${DISTNAME:C/(.).*/\1/}/${DISTNAME:C/(.*)-[0-9].*/\1/} CPAN ${PORTNAME:C/-.*//} DEBIAN pool/main/${PORTNAME:C/^((lib)?.).*$/\1/}/${PORTNAME} FARSIGHT ${PORTNAME} FESTIVAL ${PORTREVISION} GCC releases/${DISTNAME} GENTOO distfiles GIMP ${PORTNAME}/${PORTVERSION:R}/ GH ${GH_ACCOUNT}/${GH_PROJECT}/tar.gz/${GH_TAGNAME}?dummy=/ GHC ${GH_ACCOUNT}/${GH_PROJECT}/ GNOME sources/${PORTNAME}/${PORTVERSION:C/^([0-9]+\.[0-9]+).*/\1/} GNU ${PORTNAME} GNUPG ${PORTNAME} GNU_ALPHA ${PORTNAME} HORDE ${PORTNAME} LODEV ${PORTNAME} MATE ${PORTVERSION:C/^([0-9]+\.[0-9]+).*/\1/} MOZDEV ${PORTNAME:tl} NL ${PORTNAME} QT archive/qt/${PORTVERSION:R} SAMBA ${PORTNAME} SAVANNAH ${PORTNAME:tl} SF ${PORTNAME:tl}/${PORTNAME:tl}/${PORTVERSION}
<varname>USE_GITHUB</varname> If the distribution file comes from a specific commit or tag on GitHub for which there is no officially released file, there is an easy way to set the right DISTNAME and MASTER_SITES automatically. These variables are available: <varname>USE_GITHUB</varname> Description Variable Description Default GH_ACCOUNT Account name of the GitHub user hosting the project ${PORTNAME} GH_PROJECT Name of the project on GitHub ${PORTNAME} GH_TAGNAME Name of the tag to download (2.0.1, hash, ...) Using the name of a branch here is incorrect. It is also possible to use the hash of a commit id to do a snapshot. ${DISTVERSIONPREFIX}${DISTVERSION}${DISTVERSIONSUFFIX} GH_TUPLE GH_TUPLE allows putting all the GH_ACCOUNT, GH_PROJECT, and GH_TAGNAME into one variable. The format is account:project:tagname:group. It is helpful when there is more than one GitHub project from which to fetch.
Do not use GH_TUPLE for the default distribution file, as it has no default. Simple Use of <varname>USE_GITHUB</varname> While trying to make a port for version 1.2.7 of pkg from the &os; user on github, at , The Makefile would end up looking like this (slightly stripped for the example): PORTNAME= pkg PORTVERSION= 1.2.7 USE_GITHUB= yes GH_ACCOUNT= freebsd It will automatically have MASTER_SITES set to GH GHC and WRKSRC to ${WRKDIR}/pkg-1.2.7. More Complete Use of <varname>USE_GITHUB</varname> While trying to make a port for the bleeding edge version of pkg from the &os; user on github, at , the Makefile ends up looking like this (slightly stripped for the example): PORTNAME= pkg-devel PORTVERSION= 1.3.0.a.20140411 USE_GITHUB= yes GH_ACCOUNT= freebsd GH_PROJECT= pkg GH_TAGNAME= 6dbb17b It will automatically have MASTER_SITES set to GH GHC and WRKSRC to ${WRKDIR}/pkg-6dbb17b. Use of <varname>USE_GITHUB</varname> with <varname>DISTVERSIONPREFIX</varname> From time to time, GH_TAGNAME is a slight variation from DISTVERSION. For example, if the version is 1.0.2, the tag is v1.0.2. In those cases, it is possible to use DISTVERSIONPREFIX or DISTVERSIONSUFFIX: PORTNAME= foo PORTVERSION= 1.0.2 DISTVERSIONPREFIX= v USE_GITHUB= yes It will automatically set GH_TAGNAME to v1.0.2, while WRKSRC will be kept to ${WRKDIR}/foo-1.0.2. Fetching Multiple Files from GitHub The USE_GITHUB framework also supports fetching multiple distribution files from different places in GitHub. It works in a way very similar to . Multiple values are added to GH_ACCOUNT, GH_PROJECT, and GH_TAGNAME. Each different value is assigned a tag. The main value can either have no tag, or the :DEFAULT tag. A value can be omitted if it is the same as the default as listed in . GH_TUPLE can also be used when there are a lot of distribution files. It helps keep the account, project, tagname, and group information at the same place. For each tag, a ${WRKSRC_tag} helper variable is created, containing the directory into which the file has been extracted. The ${WRKSRC_tag} variables can be used to move directories around during post-extract, or add to CONFIGURE_ARGS, or whatever is needed so that the software builds correctly. Use of <varname>USE_GITHUB</varname> with Multiple Distribution Files From time to time, there is a need to fetch more than one distribution file. For example, when the upstream git repository uses submodules. This can be done easily using tags in the GH_* variables: PORTNAME= foo PORTVERSION= 1.0.2 USE_GITHUB= yes GH_ACCOUNT= bar:icons,contrib GH_PROJECT= foo-icons:icons foo-contrib:contrib GH_TAGNAME= 1.0:icons fa579bc:contrib CONFIGURE_ARGS= --with-contrib=${WRKSRC_contrib} post-extract: @${MV} ${WRKSRC_icons} ${WRKSRC}/icons This will fetch three distribution files from github. The default one comes from foo/foo and is version 1.0.2. The second one, tagged icons, comes from bar/foo-icons and is in version 1.0. The third one comes from bar/foo-contrib and uses the Git commit fa579bc. The distribution files are named foo-foo-1.0.2_GH0.tar.gz, bar-foo-icons-1.0_GH0.tar.gz, and bar-foo-contrib-fa579bc_GH0.tar.gz. All the distribution files are extracted in ${WRKDIR} in their respective subdirectories. The default file is still extracted in ${WRKSRC}, in this case, ${WRKDIR}/foo-1.0.2. Each additional distribution file is extracted in ${WRKSRC_tag}. Here, for the icons tag, it is called ${WRKSRC_icons} and it contains ${WRKDIR}/foo-icons-1.0. The file with the contrib tag is called ${WRKSRC_contrib} and contains ${WRKDIR}/foo-contrib-fa579bc. Use of <varname>USE_GITHUB</varname> with Multiple Distribution Files Using <varname>GH_TUPLE</varname> This is functionally equivalent to , but using GH_TUPLE: PORTNAME= foo PORTVERSION= 1.0.2 USE_GITHUB= yes GH_TUPLE= bar:foo-icons:1.0:icons \ bar:foo-contrib:fa579bc:contrib CONFIGURE_ARGS= --with-contrib=${WRKSRC_contrib} post-extract: @${MV} ${WRKSRC_icons} ${WRKSRC}/icons Grouping was used in the previous example with bar:icons,contrib. Some redundant information is present with GH_TUPLE because grouping is not possible. <varname>EXTRACT_SUFX</varname> If there is one distribution file, and it uses an odd suffix to indicate the compression mechanism, set EXTRACT_SUFX. For example, if the distribution file was named foo.tar.gzip instead of the more normal foo.tar.gz, write: DISTNAME= foo EXTRACT_SUFX= .tar.gzip The USES=tar[:xxx], USES=lha or USES=zip automatically set EXTRACT_SUFX to the most common archives extensions as necessary, see for more details. If neither of these are set then EXTRACT_SUFX defaults to .tar.gz. As EXTRACT_SUFX is only used in DISTFILES, only set one of them.. <varname>DISTFILES</varname> Sometimes the names of the files to be downloaded have no resemblance to the name of the port. For example, it might be called source.tar.gz or similar. In other cases the application's source code might be in several different archives, all of which must be downloaded. If this is the case, set DISTFILES to be a space separated list of all the files that must be downloaded. DISTFILES= source1.tar.gz source2.tar.gz If not explicitly set, DISTFILES defaults to ${DISTNAME}${EXTRACT_SUFX}. <varname>EXTRACT_ONLY</varname> If only some of the DISTFILES must be extracted—for example, one of them is the source code, while another is an uncompressed document—list the filenames that must be extracted in EXTRACT_ONLY. DISTFILES= source.tar.gz manual.html EXTRACT_ONLY= source.tar.gz When none of the DISTFILES need to be uncompressed, set EXTRACT_ONLY to the empty string. EXTRACT_ONLY= <varname>PATCHFILES</varname> If the port requires some additional patches that are available by FTP or HTTP, set PATCHFILES to the names of the files and PATCH_SITES to the URL of the directory that contains them (the format is the same as MASTER_SITES). If the patch is not relative to the top of the source tree (that is, WRKSRC) because it contains some extra pathnames, set PATCH_DIST_STRIP accordingly. For instance, if all the pathnames in the patch have an extra foozolix-1.0/ in front of the filenames, then set PATCH_DIST_STRIP=-p1. Do not worry if the patches are compressed; they will be decompressed automatically if the filenames end with .Z, .gz, .bz2 or .xz. If the patch is distributed with some other files, such as documentation, in a gzipped tarball, using PATCHFILES is not possible. If that is the case, add the name and the location of the patch tarball to DISTFILES and MASTER_SITES. Then, use EXTRA_PATCHES to point to those files and bsd.port.mk will automatically apply them. In particular, do not copy patch files into ${PATCHDIR}. That directory may not be writable. If there are multiple patches and they need mixed values for the strip parameter, it can be added alongside the patch name in PATCHFILES, e.g: PATCHFILES= patch1 patch2:-p1 This does not conflict with the master site grouping feature, adding a group also works: PATCHFILES= patch2:-p1:source2 The tarball will have been extracted alongside the regular source by then, so there is no need to explicitly extract it if it is a regular gzipped or compressed tarball. Take extra care not to overwrite something that already exists in that directory if extracting it manually. Also, do not forget to add a command to remove the copied patch in the pre-clean target. Multiple Distribution or Patches Files from Multiple Locations (Consider this to be a somewhat advanced topic; those new to this document may wish to skip this section at first). This section has information on the fetching mechanism known as both MASTER_SITES:n and MASTER_SITES_NN. We will refer to this mechanism as MASTER_SITES:n. A little background first. OpenBSD has a neat feature inside DISTFILES and PATCHFILES which allows files and patches to be postfixed with :n identifiers. Here, n can be both [0-9] and denote a group designation. For example: DISTFILES= alpha:0 beta:1 In OpenBSD, distribution file alpha will be associated with variable MASTER_SITES0 instead of our common MASTER_SITES and beta with MASTER_SITES1. This is a very interesting feature which can decrease that endless search for the correct download site. Just picture 2 files in DISTFILES and 20 sites in MASTER_SITES, the sites slow as hell where beta is carried by all sites in MASTER_SITES, and alpha can only be found in the 20th site. It would be such a waste to check all of them if the maintainer knew this beforehand, would it not? Not a good start for that lovely weekend! Now that you have the idea, just imagine more DISTFILES and more MASTER_SITES. Surely our distfiles survey meister would appreciate the relief to network strain that this would bring. In the next sections, information will follow on the &os; implementation of this idea. We improved a bit on OpenBSD's concept. Simplified Information This section explains how to quickly prepare fine grained fetching of multiple distribution files and patches from different sites and subdirectories. We describe here a case of simplified MASTER_SITES:n usage. This will be sufficient for most scenarios. More detailed information are available in . Some applications consist of multiple distribution files that must be downloaded from a number of different sites. For example, Ghostscript consists of the core of the program, and then a large number of driver files that are used depending on the user's printer. Some of these driver files are supplied with the core, but many others must be downloaded from a variety of different sites. To support this, each entry in DISTFILES may be followed by a colon and a tag name. Each site listed in MASTER_SITES is then followed by a colon, and the tag that indicates which distribution files are downloaded from this site. For example, consider an application with the source split in two parts, source1.tar.gz and source2.tar.gz, which must be downloaded from two different sites. The port's Makefile would include lines like . Simplified Use of <literal>MASTER_SITES:n</literal> with One File Per Site MASTER_SITES= ftp://ftp1.example.com/:source1 \ http://www.example.com/:source2 DISTFILES= source1.tar.gz:source1 \ source2.tar.gz:source2 Multiple distribution files can have the same tag. Continuing the previous example, suppose that there was a third distfile, source3.tar.gz, that is downloaded from ftp.example2.com. The Makefile would then be written like . Simplified Use of <literal>MASTER_SITES:n</literal> with More Than One File Per Site MASTER_SITES= ftp://ftp.example.com/:source1 \ http://www.example.com/:source2 DISTFILES= source1.tar.gz:source1 \ source2.tar.gz:source2 \ source3.tar.gz:source2 Detailed Information Okay, so the previous example did not reflect the new port's needs? In this section we will explain in detail how the fine grained fetching mechanism MASTER_SITES:n works and how it can be used. Elements can be postfixed with :n where n is [^:,]+, that is, n could conceptually be any alphanumeric string but we will limit it to [a-zA-Z_][0-9a-zA-Z_]+ for now. Moreover, string matching is case sensitive; that is, n is different from N. However, these words cannot be used for postfixing purposes since they yield special meaning: default, all and ALL (they are used internally in item ). Furthermore, DEFAULT is a special purpose word (check item ). Elements postfixed with :n belong to the group n, :m belong to group m and so forth. Elements without a postfix are groupless, they all belong to the special group DEFAULT. Any elements postfixed with DEFAULT, is just being redundant unless an element belongs to both DEFAULT and other groups at the same time (check item ). These examples are equivalent but the first one is preferred: MASTER_SITES= alpha MASTER_SITES= alpha:DEFAULT Groups are not exclusive, an element may belong to several different groups at the same time and a group can either have either several different elements or none at all. When an element belongs to several groups at the same time, use the comma operator (,). Instead of repeating it several times, each time with a different postfix, we can list several groups at once in a single postfix. For instance, :m,n,o marks an element that belongs to group m, n and o. All these examples are equivalent but the last one is preferred: MASTER_SITES= alpha alpha:SOME_SITE MASTER_SITES= alpha:DEFAULT alpha:SOME_SITE MASTER_SITES= alpha:SOME_SITE,DEFAULT MASTER_SITES= alpha:DEFAULT,SOME_SITE All sites within a given group are sorted according to MASTER_SORT_AWK. All groups within MASTER_SITES and PATCH_SITES are sorted as well. Group semantics can be used in any of the variables MASTER_SITES, PATCH_SITES, MASTER_SITE_SUBDIR, PATCH_SITE_SUBDIR, DISTFILES, and PATCHFILES according to this syntax: All MASTER_SITES, PATCH_SITES, MASTER_SITE_SUBDIR and PATCH_SITE_SUBDIR elements must be terminated with the forward slash / character. If any elements belong to any groups, the group postfix :n must come right after the terminator /. The MASTER_SITES:n mechanism relies on the existence of the terminator / to avoid confusing elements where a :n is a valid part of the element with occurrences where :n denotes group n. For compatibility purposes, since the / terminator was not required before in both MASTER_SITE_SUBDIR and PATCH_SITE_SUBDIR elements, if the postfix immediate preceding character is not a / then :n will be considered a valid part of the element instead of a group postfix even if an element is postfixed with :n. See both and . Detailed Use of <literal>MASTER_SITES:n</literal> in <varname>MASTER_SITE_SUBDIR</varname> MASTER_SITE_SUBDIR= old:n new/:NEW Directories within group DEFAULT -> old:n Directories within group NEW -> new Detailed Use of <literal>MASTER_SITES:n</literal> with Comma Operator, Multiple Files, Multiple Sites and Multiple Subdirectories MASTER_SITES= http://site1/%SUBDIR%/ http://site2/:DEFAULT \ http://site3/:group3 http://site4/:group4 \ http://site5/:group5 http://site6/:group6 \ http://site7/:DEFAULT,group6 \ http://site8/%SUBDIR%/:group6,group7 \ http://site9/:group8 DISTFILES= file1 file2:DEFAULT file3:group3 \ file4:group4,group5,group6 file5:grouping \ file6:group7 MASTER_SITE_SUBDIR= directory-trial:1 directory-n/:groupn \ directory-one/:group6,DEFAULT \ directory The previous example results in this fine grained fetching. Sites are listed in the exact order they will be used. file1 will be fetched from MASTER_SITE_OVERRIDE http://site1/directory-trial:1/ http://site1/directory-one/ http://site1/directory/ http://site2/ http://site7/ MASTER_SITE_BACKUP file2 will be fetched exactly as file1 since they both belong to the same group MASTER_SITE_OVERRIDE http://site1/directory-trial:1/ http://site1/directory-one/ http://site1/directory/ http://site2/ http://site7/ MASTER_SITE_BACKUP file3 will be fetched from MASTER_SITE_OVERRIDE http://site3/ MASTER_SITE_BACKUP file4 will be fetched from MASTER_SITE_OVERRIDE http://site4/ http://site5/ http://site6/ http://site7/ http://site8/directory-one/ MASTER_SITE_BACKUP file5 will be fetched from MASTER_SITE_OVERRIDE MASTER_SITE_BACKUP file6 will be fetched from MASTER_SITE_OVERRIDE http://site8/ MASTER_SITE_BACKUP How do I group one of the special macros from bsd.sites.mk, for example, SourceForge (SF)? This has been simplified as much as possible. See . Detailed Use of <literal>MASTER_SITES:n</literal> with SourceForge (<literal>SF</literal>) MASTER_SITES= http://site1/ SF/something/1.0:sourceforge,TEST DISTFILES= something.tar.gz:sourceforge something.tar.gz will be fetched from all sites within SourceForge. How do I use this with PATCH*? All examples were done with MASTER* but they work exactly the same for PATCH* ones as can be seen in . Simplified Use of <literal>MASTER_SITES:n</literal> with <varname>PATCH_SITES</varname> PATCH_SITES= http://site1/ http://site2/:test PATCHFILES= patch1:test What Does Change for Ports? What Does Not? All current ports remain the same. The MASTER_SITES:n feature code is only activated if there are elements postfixed with :n like elements according to the aforementioned syntax rules, especially as shown in item . The port targets remain the same: checksum, makesum, patch, configure, build, etc. With the obvious exceptions of do-fetch, fetch-list, master-sites and patch-sites. do-fetch: deploys the new grouping postfixed DISTFILES and PATCHFILES with their matching group elements within both MASTER_SITES and PATCH_SITES which use matching group elements within both MASTER_SITE_SUBDIR and PATCH_SITE_SUBDIR. Check . fetch-list: works like old fetch-list with the exception that it groups just like do-fetch. master-sites and patch-sites: (incompatible with older versions) only return the elements of group DEFAULT; in fact, they execute targets master-sites-default and patch-sites-default respectively. Furthermore, using target either master-sites-all or patch-sites-all is preferred to directly checking either MASTER_SITES or PATCH_SITES. Also, directly checking is not guaranteed to work in any future versions. Check item for more information on these new port targets. New port targets There are master-sites-n and patch-sites-n targets which will list the elements of the respective group n within MASTER_SITES and PATCH_SITES respectively. For instance, both master-sites-DEFAULT and patch-sites-DEFAULT will return the elements of group DEFAULT, master-sites-test and patch-sites-test of group test, and thereon. There are new targets master-sites-all and patch-sites-all which do the work of the old master-sites and patch-sites ones. They return the elements of all groups as if they all belonged to the same group with the caveat that it lists as many MASTER_SITE_BACKUP and MASTER_SITE_OVERRIDE as there are groups defined within either DISTFILES or PATCHFILES; respectively for master-sites-all and patch-sites-all. <varname>DIST_SUBDIR</varname> Do not let the port clutter /usr/ports/distfiles. If the port requires a lot of files to be fetched, or contains a file that has a name that might conflict with other ports (for example, Makefile), set DIST_SUBDIR to the name of the port (${PORTNAME} or ${PKGNAMEPREFIX}${PORTNAME} are fine). This will change DISTDIR from the default /usr/ports/distfiles to /usr/ports/distfiles/${DIST_SUBDIR}, and in effect puts everything that is required for the port into that subdirectory. It will also look at the subdirectory with the same name on the backup master site at ftp.FreeBSD.org. (Setting DISTDIR explicitly in Makefile will not accomplish this, so please use DIST_SUBDIR.) This does not affect MASTER_SITES defined in the Makefile. <varname>ALWAYS_KEEP_DISTFILES</varname> If the port uses binary distfiles and has a license that requires that the source code is provided with packages distributed in binary form, like GPL, ALWAYS_KEEP_DISTFILES will instruct the &os; build cluster to keep a copy of the files specified in DISTFILES. Users of these ports will generally not need these files, so it is a good idea to only add the source distfiles to DISTFILES when PACKAGE_BUILDING is defined. Use of <varname>ALWAYS_KEEP_DISTFILES</varname> .if defined(PACKAGE_BUILDING) DISTFILES+= foo.tar.gz ALWAYS_KEEP_DISTFILES= yes .endif When adding extra files to DISTFILES, make sure to also add them to distinfo. Also, the additional files will normally be extracted into WRKDIR as well, which for some ports may lead to undesirable side effects and require special handling. <varname>MAINTAINER</varname> Set your mail-address here. Please. :-) Only a single address without the comment part is allowed as a MAINTAINER value. The format used is user@hostname.domain. Please do not include any descriptive text such as a real name in this entry. That merely confuses the Ports infrastructure and most tools using it. The maintainer is responsible for keeping the port up to date and making sure that it works correctly. For a detailed description of the responsibilities of a port maintainer, refer to The challenge for port maintainers. A maintainer volunteers to keep a port in good working order. Maintainers have the primary responsibility for their ports, but not exclusive ownership. Ports exist for the benefit of the community and, in reality, belong to the community. What this means is that people other than the maintainer can make changes to a port. Large changes to the Ports Collection might require changes to many ports. The &os; Ports Management Team or members of other teams might modify ports to fix dependency issues or other problems, like a version bump for a shared library update. Some types of fixes have blanket approval from the &a.portmgr;, allowing any committer to fix those categories of problems on any port. These fixes do not need approval from the maintainer. Blanket approval does not apply to ports that are maintained by teams like autotools@FreeBSD.org, x11@FreeBSD.org, gnome@FreeBSD.org, or kde@FreeBSD.org. These teams use external repositories and can have work that would conflict with changes that would normally fall under blanket approval. Blanket approval for most ports applies to these types of fixes: Most infrastructure changes to a port (that is, modernizing, but not changing the functionality). For example, converting to staging, USE_GMAKE to USES=gmake, the new LIB_DEPENDS format... Trivial and tested build and runtime fixes. Other changes to the port will be sent to the maintainer for review and approval before being committed. If the maintainer does not respond to an update request after two weeks (excluding major public holidays), then that is considered a maintainer timeout, and the update may be made without explicit maintainer approval. If the maintainer does not respond within three months, or if there have been three consecutive timeouts, then that maintainer is considered absent without leave, and can be replaced as the maintainer of the particular port in question. Exceptions to this are anything maintained by the &a.portmgr;, or the &a.security-officer;. No unauthorized commits may ever be made to ports maintained by those groups. We reserve the right to modify the maintainer's submission to better match existing policies and style of the Ports Collection without explicit blessing from the submitter or the maintainer. Also, large infrastructural changes can result in a port being modified without the maintainer's consent. These kinds of changes will never affect the port's functionality. The &a.portmgr; reserves the right to revoke or override anyone's maintainership for any reason, and the &a.security-officer; reserves the right to revoke or override maintainership for security reasons. <varname>COMMENT</varname> This is a one-line description of the port. Please respect these rules: Try to keep the COMMENT value at no longer than 70 characters, as this line will be used by pkg info (see &man.pkg-info.8;) to display a one-line summary of the port; Do not include the package name (or version number of the software); The comment must begin with a capital and end without a period; Do not start with an indefinite article (that is, A or An); Names are capitalized (for example, Apache, JavaScript, Perl); For lists of words, use the Oxford comma (for example, green, red, and blue); Spell check the text. Here is an example: COMMENT= Cat chasing a mouse all over the screen The COMMENT variable immediately follows the MAINTAINER variable in the Makefile. <varname>PORTSCOUT</varname> Portscout is an automated distfile check utility for the &os; Ports Collection, described in detail in . PORTSCOUT defines special conditions within which the Portscout distfile scanner is restricted. Situations where PORTSCOUT is set include: When distfiles have to be ignored, whether for specific versions, or specific minor revisions. For example, to exclude version 8.2 from distfile version checks because it is known to be broken, add: PORTSCOUT= ignore:8.2 When specific versions or specific major and minor revisions of a distfile must be checked. For example, if only version 0.6.4 must be monitored because newer versions have compatibility issues with &os;, add: PORTSCOUT= limit:^0\.6\.4 When URLs listing the available versions differ from the download URLs. For example, to limit distfile version checks to the download page for the databases/pgtune port, add: PORTSCOUT= site:http://pgfoundry.org/frs/?group_id=1000416 Dependencies Many ports depend on other ports. This is a very convenient feature of most Unix-like operating systems, including &os;. Multiple ports can share a common dependency, rather than bundling that dependency with every port or package that needs it. There are seven variables that can be used to ensure that all the required bits will be on the user's machine. There are also some pre-supported dependency variables for common cases, plus a few more to control the behavior of dependencies. <varname>LIB_DEPENDS</varname> This variable specifies the shared libraries this port depends on. It is a list of lib:dir tuples where lib is the name of the shared library, dir is the directory in which to find it in case it is not available. For example, LIB_DEPENDS= libjpeg.so:${PORTSDIR}/graphics/jpeg will check for a shared jpeg library with any version, and descend into the graphics/jpeg subdirectory of the ports tree to build and install it if it is not found. The dependency is checked twice, once from within the build target and then from within the install target. Also, the name of the dependency is put into the package so that pkg install (see &man.pkg-install.8;) will automatically install it if it is not on the user's system. <varname>RUN_DEPENDS</varname> This variable specifies executables or files this port depends on during run-time. It is a list of path:dir:target tuples where path is the name of the executable or file, dir is the directory in which to find it in case it is not available, and target is the target to call in that directory. If path starts with a slash (/), it is treated as a file and its existence is tested with test -e; otherwise, it is assumed to be an executable, and which -s is used to determine if the program exists in the search path. For example, RUN_DEPENDS= ${LOCALBASE}/news/bin/innd:${PORTSDIR}/news/inn \ xmlcatmgr:${PORTSDIR}/textproc/xmlcatmgr will check if the file or directory /usr/local/news/bin/innd exists, and build and install it from the news/inn subdirectory of the ports tree if it is not found. It will also see if an executable called xmlcatmgr is in the search path, and descend into textproc/xmlcatmgr to build and install it if it is not found. In this case, innd is actually an executable; if an executable is in a place that is not expected to be in the search path, use the full pathname. The official search PATH used on the ports build cluster is /sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin The dependency is checked from within the install target. Also, the name of the dependency is put into the package so that pkg install (see &man.pkg-install.8;) will automatically install it if it is not on the user's system. The target part can be omitted if it is the same as DEPENDS_TARGET. A quite common situation is when RUN_DEPENDS is literally the same as BUILD_DEPENDS, especially if ported software is written in a scripted language or if it requires the same build and run-time environment. In this case, it is both tempting and intuitive to directly assign one to the other: RUN_DEPENDS= ${BUILD_DEPENDS} However, such assignment can pollute run-time dependencies with entries not defined in the port's original BUILD_DEPENDS. This happens because of &man.make.1;'s lazy evaluation of variable assignment. Consider a Makefile with USE_*, which are processed by ports/Mk/bsd.*.mk to augment initial build dependencies. For example, USES= gmake adds devel/gmake to BUILD_DEPENDS. To prevent such additional dependencies from polluting RUN_DEPENDS, create another variable with the current content of BUILD_DEPENDS and assign it to both BUILD_DEPENDS and RUN_DEPENDS: MY_DEPENDS= some:${PORTSDIR}/devel/some \ other:${PORTSDIR}/lang/other BUILD_DEPENDS= ${MY_DEPENDS} RUN_DEPENDS= ${MY_DEPENDS} Do not use := to assign BUILD_DEPENDS to RUN_DEPENDS or vice-versa. All variables are expanded immediately, which is exactly the wrong thing to do and almost always a failure. <varname>BUILD_DEPENDS</varname> This variable specifies executables or files this port requires to build. Like RUN_DEPENDS, it is a list of path:dir:target tuples. For example, BUILD_DEPENDS= unzip:${PORTSDIR}/archivers/unzip will check for an executable called unzip, and descend into the archivers/unzip subdirectory of the ports tree to build and install it if it is not found. build here means everything from extraction to compilation. The dependency is checked from within the extract target. The target part can be omitted if it is the same as DEPENDS_TARGET <varname>FETCH_DEPENDS</varname> This variable specifies executables or files this port requires to fetch. Like the previous two, it is a list of path:dir:target tuples. For example, FETCH_DEPENDS= ncftp2:${PORTSDIR}/net/ncftp2 will check for an executable called ncftp2, and descend into the net/ncftp2 subdirectory of the ports tree to build and install it if it is not found. The dependency is checked from within the fetch target. The target part can be omitted if it is the same as DEPENDS_TARGET. <varname>EXTRACT_DEPENDS</varname> This variable specifies executables or files this port requires for extraction. Like the previous, it is a list of path:dir:target tuples. For example, EXTRACT_DEPENDS= unzip:${PORTSDIR}/archivers/unzip will check for an executable called unzip, and descend into the archivers/unzip subdirectory of the ports tree to build and install it if it is not found. The dependency is checked from within the extract target. The target part can be omitted if it is the same as DEPENDS_TARGET. Use this variable only if the extraction does not already work (the default assumes tar) and cannot be made to work using USES=tar, USES=lha or USES=zip described in . <varname>PATCH_DEPENDS</varname> This variable specifies executables or files this port requires to patch. Like the previous, it is a list of path:dir:target tuples. For example, PATCH_DEPENDS= ${NONEXISTENT}:${PORTSDIR}/java/jfc:extract will descend into the java/jfc subdirectory of the ports tree to extract it. The dependency is checked from within the patch target. The target part can be omitted if it is the same as DEPENDS_TARGET. <varname>USES</varname> Parameters can be added to define different features and dependencies used by the port. They are specified by adding this line to the Makefile: USES= feature[:arguments] For the complete list of values, please see . USES cannot be assigned after inclusion of bsd.port.pre.mk. <varname>USE_<replaceable>*</replaceable></varname> Several variables exist to define common dependencies shared by many ports. Their use is optional, but helps to reduce the verbosity of the port Makefiles. Each of them is styled as USE_*. These variables may be used only in the port Makefiles and ports/Mk/bsd.*.mk. They are not meant for user-settable options — use PORT_OPTIONS for that purpose. It is always incorrect to set any USE_* in /etc/make.conf. For instance, setting USE_GCC=X.Y (where X.Y is version number) would add a dependency on gccXY for every port, including lang/gccXY itself! <varname>USE_<replaceable>*</replaceable></varname> Variable Means USE_GCC The port requires GCC (gcc or g++) to build. Some ports need any GCC version, some require modern, recent versions. It is typically set to any (in this case, GCC from base would be used on versions of &os; that still have it, or lang/gcc port would be installed when default C/C++ compiler is Clang); or yes (means always use stable, modern GCC from lang/gcc port). The exact version can also be specified, with a value such as 4.7. The minimal required version can be specified as 4.6+. The GCC from the base system is used when it satisfies the requested version, otherwise an appropriate compiler is built from the port, and CC and CXX are adjusted accordingly.
Variables related to gmake and configure are described in , while autoconf, automake and libtool are described in . Perl related variables are described in . X11 variables are listed in . deals with GNOME and with KDE related variables. documents Java variables, while contains information on Apache, PHP and PEAR modules. Python is discussed in , while Ruby in . provides variables used for SDL applications and finally, contains information on Xfce. Minimal Version of a Dependency A minimal version of a dependency can be specified in any *_DEPENDS except LIB_DEPENDS using this syntax: p5-Spiffy>=0.26:${PORTSDIR}/devel/p5-Spiffy The first field contains a dependent package name, which must match the entry in the package database, a comparison sign, and a package version. The dependency is satisfied if p5-Spiffy-0.26 or newer is installed on the machine. Notes on Dependencies As mentioned above, the default target to call when a dependency is required is DEPENDS_TARGET. It defaults to install. This is a user variable; it is never defined in a port's Makefile. If the port needs a special way to handle a dependency, use the :target part of *_DEPENDS instead of redefining DEPENDS_TARGET. When running make clean, the port dependencies are automatically cleaned too. If this is not desirable, define NOCLEANDEPENDS in the environment. This may be particularly desirable if the port has something that takes a long time to rebuild in its dependency list, such as KDE, GNOME or Mozilla. To depend on another port unconditionally, use the variable ${NONEXISTENT} as the first field of BUILD_DEPENDS or RUN_DEPENDS. Use this only when the source of the other port is needed. Compilation time can be saved by specifying the target too. For instance BUILD_DEPENDS= ${NONEXISTENT}:${PORTSDIR}/graphics/jpeg:extract will always descend to the jpeg port and extract it. Circular Dependencies Are Fatal Do not introduce any circular dependencies into the ports tree! The ports building technology does not tolerate circular dependencies. If one is introduced, someone, somewhere in the world, will have their &os; installation broken almost immediately, with many others quickly to follow. These can really be hard to detect. If in doubt, before making that change, make sure to run: cd /usr/ports; make index. That process can be quite slow on older machines, but it may be able to save a large number of people, including yourself, a lot of grief in the process. Problems Caused by Automatic Dependencies Dependencies must be declared either explicitly or by using the OPTIONS framework. Using other methods like automatic detection complicates indexing, which causes problems for port and package management. Wrong Declaration of an Optional Dependency .include <bsd.port.pre.mk> .if exists(${LOCALBASE}/bin/foo) LIB_DEPENDS= libbar.so:${PORTSDIR}/foo/bar .endif The problem with trying to automatically add dependencies is that files and settings outside an individual port can change at any time. For example: an index is built, then a batch of ports are installed. But one of the ports installs the tested file. The index is now incorrect, because an installed port unexpectedly has a new dependency. The index may still be wrong even after rebuilding if other ports also determine their need for dependencies based on the existence of other files. Correct Declaration of an Optional Dependency OPTIONS_DEFINE= BAR BAR_DESC= Calling cellphones via bar BAR_LIB_DEPENDS= libbar.so:${PORTSDIR}/foo/bar Testing option variables is the correct method. It will not cause inconsistencies in the index of a batch of ports, provided the options were defined prior to the index build. Simple scripts can then be used to automate the building, installation, and updating of these ports and their packages. <varname>USE_<replaceable>*</replaceable></varname> and <varname>WANT_<replaceable>*</replaceable></varname> USE_* are set by the port maintainer to define software on which this port depends. A port that needs Firefox would set USE_FIREFOX= yes Some USE_* can accept version numbers or other parameters. For example, a port that requires Apache 2.2 would set USE_APACHE= 22 For more control over dependencies in some cases, WANT_* are available to more precisely specify what is needed. For example, consider the mail/squirrelmail port. This port needs some PHP modules, which are listed in USE_PHP: USE_PHP= session mhash gettext mbstring pcre openssl xml Those modules may be available in CLI or web versions, so the web version is selected with WANT_*: WANT_PHP_WEB= yes Available USE_* and WANT_* are defined in the files in /usr/ports/Mk. <varname>MASTERDIR</varname> If the port needs to build slightly different versions of packages by having a variable (for instance, resolution, or paper size) take different values, create one subdirectory per package to make it easier for users to see what to do, but try to share as many files as possible between ports. Typically, by using variables cleverly, only a very short Makefile is needed in all but one of the directories. In the sole Makefile, use MASTERDIR to specify the directory where the rest of the files are. Also, use a variable as part of PKGNAMESUFFIX so the packages will have different names. This will be best demonstrated by an example. This is part of japanese/xdvi300/Makefile; PORTNAME= xdvi PORTVERSION= 17 PKGNAMEPREFIX= ja- PKGNAMESUFFIX= ${RESOLUTION} # default RESOLUTION?= 300 .if ${RESOLUTION} != 118 && ${RESOLUTION} != 240 && \ ${RESOLUTION} != 300 && ${RESOLUTION} != 400 pre-everything:: @${ECHO_MSG} "Error: invalid value for RESOLUTION: \"${RESOLUTION}\"" @${ECHO_MSG} "Possible values are: 118, 240, 300 (default) and 400." @${FALSE} .endif japanese/xdvi300 also has all the regular patches, package files, etc. Running make there, it will take the default value for the resolution (300) and build the port normally. As for other resolutions, this is the entire xdvi118/Makefile: RESOLUTION= 118 MASTERDIR= ${.CURDIR}/../xdvi300 .include "${MASTERDIR}/Makefile" (xdvi240/Makefile and xdvi400/Makefile are similar). MASTERDIR definition tells bsd.port.mk that the regular set of subdirectories like FILESDIR and SCRIPTDIR are to be found under xdvi300. The RESOLUTION=118 line will override the RESOLUTION=300 line in xdvi300/Makefile and the port will be built with resolution set to 118. Man Pages If the port anchors its man tree somewhere other than PREFIX, use MANDIRS to specify those directories. Note that the files corresponding to manual pages must be placed in pkg-plist along with the rest of the files. The purpose of MANDIRS is to enable automatic compression of manual pages, therefore the file names are suffixed with .gz. Info Files If the package needs to install GNU info files, list them in INFO (without the trailing .info), one entry per document. These files are assumed to be installed to PREFIX/INFO_PATH. Change INFO_PATH if the package uses a different location. However, this is not recommended. These entries contain just the path relative to PREFIX/INFO_PATH. For example, lang/gcc34 installs info files to PREFIX/INFO_PATH/gcc34, and INFO will be something like this: INFO= gcc34/cpp gcc34/cppinternals gcc34/g77 ... Appropriate installation/de-installation code will be automatically added to the temporary pkg-plist before package registration. Makefile Options Many applications can be built with optional or differing configurations. Examples include choice of natural (human) language, GUI versus command-line, or type of database to support. Users may need a different configuration than the default, so the ports system provides hooks the port author can use to control which variant will be built. Supporting these options properly will make users happy, and effectively provide two or more ports for the price of one. <varname>OPTIONS</varname> Background OPTIONS_* give the user installing the port a dialog showing the available options, and then saves those options to ${PORT_DBDIR}/${OPTIONS_NAME}/options. The next time the port is built, the options are reused. PORT_DBDIR defaults to /var/db/ports. OPTIONS_NAME is to the port origin with an underscore as the space separator, for example, for dns/bind99 it will be dns_bind99. When the user runs make config (or runs make build for the first time), the framework checks for ${PORT_DBDIR}/${OPTIONS_NAME}/options. If that file does not exist, the values of OPTIONS_* are used, and a dialog box is displayed where the options can be enabled or disabled. Then options is saved and the configured variables are used when building the port. If a new version of the port adds new OPTIONS, the dialog will be presented to the user with the saved values of old OPTIONS prefilled. make showconfig shows the saved configuration. Use make rmconfig to remove the saved configuration. Syntax OPTIONS_DEFINE contains a list of OPTIONS to be used. These are independent of each other and are not grouped: OPTIONS_DEFINE= OPT1 OPT2 Once defined, OPTIONS are described (optional, but strongly recommended): OPT1_DESC= Describe OPT1 OPT2_DESC= Describe OPT2 OPT3_DESC= Describe OPT3 OPT4_DESC= Describe OPT4 OPT5_DESC= Describe OPT5 OPT6_DESC= Describe OPT6 ports/Mk/bsd.options.desc.mk has descriptions for many common OPTIONS. While often useful, override them if the description is insufficient for the port. When describing options, view it from the perspective of the user: What functionality does it change? and Why would I want to enable this? Do not just repeat the name. For example, describing the NLS option as include NLS support does not help the user, who can already see the option name but may not know what it means. Describing it as Native Language Support via gettext utilities is much more helpful. Option names are always in all uppercase. They cannot use mixed case or lowercase. OPTIONS can be grouped as radio choices, where only one choice from each group is allowed: OPTIONS_SINGLE= SG1 OPTIONS_SINGLE_SG1= OPT3 OPT4 There must be one of each OPTIONS_SINGLE group selected at all times for the options to be valid. One option of each group must be added to OPTIONS_DEFAULT. OPTIONS can be grouped as radio choices, where none or only one choice from each group is allowed: OPTIONS_RADIO= RG1 OPTIONS_RADIO_RG1= OPT7 OPT8 OPTIONS can also be grouped as multiple-choice lists, where at least one option must be enabled: OPTIONS_MULTI= MG1 OPTIONS_MULTI_MG1= OPT5 OPT6 OPTIONS can also be grouped as multiple-choice lists, where none or any option can be enabled: OPTIONS_GROUP= GG1 OPTIONS_GROUP_GG1= OPT9 OPT10 OPTIONS are unset by default, unless they are listed in OPTIONS_DEFAULT: OPTIONS_DEFAULT= OPT1 OPT3 OPT6 OPTIONS definitions must appear before the inclusion of bsd.port.options.mk. PORT_OPTIONS values can only be tested after the inclusion of bsd.port.options.mk. Inclusion of bsd.port.pre.mk can be used instead, too, and is still widely used in ports written before the introduction of bsd.port.options.mk. But be aware that some variables will not work as expected after the inclusion of bsd.port.pre.mk, typically some USE_* flags. Simple Use of <varname>OPTIONS</varname> OPTIONS_DEFINE= FOO BAR FOO_DESC= Option foo support BAR_DESC= Feature bar support OPTIONS_DEFAULT=FOO # Will add --with-foo / --without-foo FOO_CONFIGURE_WITH= foo BAR_RUN_DEPENDS= bar:${PORTSDIR}/bar/bar .include <bsd.port.mk> Check for Unset Port <varname>OPTIONS</varname> .if ! ${PORT_OPTIONS:MEXAMPLES} CONFIGURE_ARGS+=--without-examples .endif The form shown above is discouraged. The preferred method is using a configure knob to really enable and disable the feature to match the option: # Will add --with-examples / --without-examples EXAMPLES_CONFIGURE_WITH= examples Practical Use of <varname>OPTIONS</varname> OPTIONS_DEFINE= EXAMPLES OPTIONS_SINGLE= BACKEND OPTIONS_SINGLE_BACKEND= MYSQL PGSQL BDB OPTIONS_MULTI= AUTH OPTIONS_MULTI_AUTH= LDAP PAM SSL EXAMPLES_DESC= Install extra examples MYSQL_DESC= Use MySQL as backend PGSQL_DESC= Use PostgreSQL as backend BDB_DESC= Use Berkeley DB as backend LDAP_DESC= Build with LDAP authentication support PAM_DESC= Build with PAM support SSL_DESC= Build with OpenSSL support OPTIONS_DEFAULT= PGSQL LDAP SSL # Will add USE_PGSQL=yes PGSQL_USE= pgsql=yes # Will add --enable-postgres / --disable-postgres PGSQL_CONFIGURE_ENABLE= postgres ICU_LIB_DEPENDS= libicuuc.so:${PORTSDIR}/devel/icu # Will add --with-examples / --without-examples EXAMPLES_CONFIGURE_WITH= examples # Check other OPTIONS .include <bsd.port.mk> Default Options These options are always on by default. DOCS — build and install documentation. NLS — Native Language Support. EXAMPLES — build and install examples. IPV6 — IPv6 protocol support. There is no need to add these to OPTIONS_DEFAULT. To have them active, and show up in the options selection dialog, however, they must be added to OPTIONS_DEFINE. Feature Auto-Activation When using a GNU configure script, keep an eye on which optional features are activated by auto-detection. Explicitly disable optional features that are not needed by adding --without-xxx or --disable-xxx in CONFIGURE_ARGS. Wrong Handling of an Option .if ${PORT_OPTIONS:MFOO} LIB_DEPENDS+= libfoo.so:${PORTSDIR}/devel/foo CONFIGURE_ARGS+= --enable-foo .endif In the example above, imagine a library libfoo is installed on the system. The user does not want this application to use libfoo, so he toggled the option off in the make config dialog. But the application's configure script detects the library present in the system and includes its support in the resulting executable. Now when the user decides to remove libfoo from the system, the ports system does not protest (no dependency on libfoo was recorded) but the application breaks. Correct Handling of an Option FOO_LIB_DEPENDS= libfoo.so:${PORTSDIR}/devel/foo # Will add --enable-foo / --disable-foo FOO_CONFIGURE_ENABLE= foo Under some circumstances, the shorthand conditional syntax can cause problems with complex constructs. The errors are usually Malformed conditional, an alternative syntax can be used. .if !empty(VARIABLE:MVALUE) as an alternative to .if ${VARIABLE:MVALUE} Options Helpers There are some macros to help simplify conditional values which differ based on the options set. <varname>OPTIONS_SUB</varname> If OPTIONS_SUB is set to yes then each of the options added to OPTIONS_DEFINE will be added to PLIST_SUB and SUB_LIST, for example: OPTIONS_DEFINE= OPT1 OPTIONS_SUB= yes is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} PLIST_SUB+= OPT1="" NO_OPT1="@comment " SUB_LIST+= OPT1="" NO_OPT1="@comment " .else PLIST_SUB+= OPT1="@comment " NO_OPT1="" SUB_LIST+= OPT1="@comment " NO_OPT1="" .endif The value of OPTIONS_SUB is ignored. Setting it to any value will add PLIST_SUB and SUB_LIST entries for all options. <varname><replaceable>OPT</replaceable>_USE</varname> and <varname><replaceable>OPT</replaceable>_USE_OFF</varname> When option OPT is selected, for each key=value pair in OPT_USE, value is appended to the corresponding USE_KEY. If value has spaces in it, replace them with commas and they will be changed back to spaces during processing. OPT_USE_OFF works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_USE= mysql=yes xorg=x11,xextproto,xext,xrandr OPT1_USE_OFF= openssl=yes is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} USE_MYSQL= yes USE_XORG= x11 xextproto xext xrandr .else USE_OPENSSL= yes .endif <varname><replaceable>OPT</replaceable>_CONFIGURE_ENABLE</varname> When option OPT is selected, for each entry in OPT_CONFIGURE_ENABLE then --enable-entry is appended to CONFIGURE_ARGS. When option OPT is not selected, --disable-entry is appended to CONFIGURE_ARGS. An optional argument can be specified with an = symbol. This argument is only appended to the --enable-entry configure option. For example: OPTIONS_DEFINE= OPT1 OPT2 OPT1_CONFIGURE_ENABLE= test1 test2 OPT2_CONFIGURE_ENABLE= test2=exhaustive is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} CONFIGURE_ARGS+= --enable-test1 --enable-test2 .else CONFIGURE_ARGS+= --disable-test1 --disable-test2 .endif .if ${PORT_OPTIONS:MOPT2} CONFIGURE_ARGS+= --enable-test2=exhaustive .else CONFIGURE_ARGS+= --disable-test2 .endif <varname><replaceable>OPT</replaceable>_CONFIGURE_WITH</varname> When option OPT is selected, for each entry in OPT_CONFIGURE_WITH then --with-entry is appended to CONFIGURE_ARGS. When option OPT is not selected, --without-entry is appended to CONFIGURE_ARGS. An optional argument can be specified with an = symbol. This argument is only appended to the --with-entry configure option. For example: OPTIONS_DEFINE= OPT1 OPT2 OPT1_CONFIGURE_WITH= test1 OPT2_CONFIGURE_WITH= test2=exhaustive is equivalent to: OPTIONS_DEFINE= OPT1 OPT2 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} CONFIGURE_ARGS+= --with-test1 .else CONFIGURE_ARGS+= --without-test1 .endif .if ${PORT_OPTIONS:MOPT2} CONFIGURE_ARGS+= --with-test2=exhaustive .else CONFIGURE_ARGS+= --without-test2 .endif <varname><replaceable>OPT</replaceable>_CONFIGURE_ON</varname> and <varname><replaceable>OPT</replaceable>_CONFIGURE_OFF</varname> When option OPT is selected, the value of OPT_CONFIGURE_ON, if defined, is appended to CONFIGURE_ARGS. OPT_CONFIGURE_OFF works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_CONFIGURE_ON= --add-test OPT1_CONFIGURE_OFF= --no-test is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} CONFIGURE_ARGS+= --add-test .else CONFIGURE_ARGS+= --no-test .endif <varname><replaceable>OPT</replaceable>_CMAKE_ON</varname> and <varname><replaceable>OPT</replaceable>_CMAKE_OFF</varname> When option OPT is selected, the value of OPT_CMAKE_ON, if defined, is appended to CMAKE_ARGS. OPT_CMAKE_OFF works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_CMAKE_ON= -DTEST:BOOL=true OPT1_CMAKE_OFF= -DOPTIMIZE:BOOL=true is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} CMAKE_ARGS+= -DTEST:BOOL=true .else CMAKE_ARGS+= -DOPTIMIZE:BOOL=true .endif <varname><replaceable>OPT</replaceable>_QMAKE_ON</varname> and <varname><replaceable>OPT</replaceable>_QMAKE_OFF</varname> When option OPT is selected, the value of OPT_QMAKE_ON, if defined, is appended to QMAKE_ARGS. OPT_QMAKE_OFF works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_QMAKE_ON= -DTEST:BOOL=true OPT1_QMAKE_OFF= -DPRODUCTION:BOOL=true is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} QMAKE_ARGS+= -DTEST:BOOL=true .else QMAKE_ARGS+= -DPRODUCTION:BOOL=true .endif <varname><replaceable>OPT</replaceable>_IMPLIES</varname> Provides a way to add dependencies between options. When OPT is selected, all the options listed in this variable will be selected too. Using the OPT_CONFIGURE_ENABLE described earlier to illustrate: OPTIONS_DEFINE= OPT1 OPT2 OPT1_IMPLIES= OPT2 OPT1_CONFIGURE_ENABLE= opt1 OPT2_CONFIGURE_ENABLE= opt2 Is equivalent to: OPTIONS_DEFINE= OPT1 OPT2 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} CONFIGURE_ARGS+= --enable-opt1 .else CONFIGURE_ARGS+= --disable-opt1 .endif .if ${PORT_OPTIONS:MOPT2} || ${PORT_OPTIONS:MOPT1} CONFIGURE_ARGS+= --enable-opt2 .else CONFIGURE_ARGS+= --disable-opt2 .endif Simple Use of <varname><replaceable>OPT</replaceable>_IMPLIES</varname> This port has a X11 option, and a GNOME option that needs the X11 option to be selected to build. OPTIONS_DEFINE= X11 GNOME OPTIONS_DEFAULT= X11 X11_USE= xorg=xi,xextproto GNOME_USE= gnome=gtk30 GNOME_IMPLIES= X11 <varname><replaceable>OPT</replaceable>_PREVENTS</varname> and <varname><replaceable>OPT</replaceable>_PREVENTS_MSG</varname> Provides a way to add conflicts between options. When OPT is selected, all the options listed in this variable must be un-selected. If OPT_PREVENTS_MSG is also selected, its content will be shown, explaining why they conflict. For example: OPTIONS_DEFINE= OPT1 OPT2 OPT1_PREVENTS= OPT2 OPT1_PREVENTS_MSG= OPT1 and OPT2 enable conflicting options Is roughly equivalent to: OPTIONS_DEFINE= OPT1 OPT2 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT2} || ${PORT_OPTIONS:MOPT1} BROKEN= Option OPT1 conflicts with OPT2 (select only one) .endif The only difference is that the first one will write an error after running make config, suggesting changing the selected options. Simple Use of <varname><replaceable>OPT</replaceable>_PREVENTS</varname> This port has X509 and SCTP options. Both options add patches, but the patches conflict with each other, so they cannot be selected at the same time. OPTIONS_DEFINE= X509 SCTP SCTP_PATCHFILES= ${PORTNAME}-6.8p1-sctp-2573.patch.gz:-p1 SCTP_CONFIGURE_WITH= sctp X509_PATCH_SITES= http://www.roumenpetrov.info/openssh/x509/:x509 X509_PATCHFILES= ${PORTNAME}-7.0p1+x509-8.5.diff.gz:-p1:x509 X509_PREVENTS= SCTP X509_PREVENTS_MSG= X509 and SCTP patches conflict <varname><replaceable>OPT</replaceable>_VARS</varname> and <varname><replaceable>OPT</replaceable>_VARS_OFF</varname> Provides a generic way to set and append to variables. Before using OPT_VARS and OPT_VARS_OFF, see if there is already a more specific helper available in . When option OPT is selected, and OPT_VARS defined, key=value and key+=value pairs are evaluated from OPT_VARS. An = cause the existing value of KEY to be overwritten, an += appends to the value. OPT_VARS_OFF works the same way, but when OPT is not selected. OPTIONS_DEFINE= OPT1 OPT2 OPT3 OPT1_VARS= also_build+=bin1 OPT2_VARS= also_build+=bin2 OPT3_VARS= bin3_build=yes OPT3_VARS_OFF= bin3_build=no MAKE_ARGS= ALSO_BUILD="${ALSO_BUILD}" BIN3_BUILD="${BIN3_BUILD}" is equivalent to: OPTIONS_DEFINE= OPT1 OPT2 MAKE_ARGS= ALSO_BUILD="${ALSO_BUILD}" BIN3_BUILD="${BIN3_BUILD}" .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} ALSO_BUILD+= bin1 .endif .if ${PORT_OPTIONS:MOPT2} ALSO_BUILD+= bin2 .endif .if ${PORT_OPTIONS:MOPT2} BIN3_BUILD= yes .else BIN3_BUILD= no .endif Values containing whitespace must be enclosed in quotes: OPT_VARS= foo="bar baz" This is due to the way &man.make.1; variable expansion deals with whitespace. When OPT_VARS= foo=bar baz is expanded, the variable ends up containing two strings, foo=bar and baz. But the submitter probably intended there to be only one string, foo=bar baz. Quoting the value prevents whitespace from being used as a delimiter. Dependencies, <varname><replaceable>OPT</replaceable>_<replaceable>DEPTYPE</replaceable></varname> and <varname><replaceable>OPT</replaceable>_<replaceable>DEPTYPE</replaceable>_OFF</varname> For any of these dependency types: PKG_DEPENDS EXTRACT_DEPENDS PATCH_DEPENDS FETCH_DEPENDS BUILD_DEPENDS LIB_DEPENDS RUN_DEPENDS When option OPT is selected, the value of OPT_DEPTYPE, if defined, is appended to DEPTYPE. OPT_DEPTYPE_OFF works the same, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_LIB_DEPENDS= liba.so:${PORTSDIR}/devel/a OPT1_LIB_DEPENDS_OFF= libb.so:${PORTSDIR}/devel/b is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} LIB_DEPENDS+= liba.so:${PORTSDIR}/devel/a .else LIB_DEPENDS+= libb.so:${PORTSDIR}/devel/b .endif Generic Variables Replacement, <varname><replaceable>OPT</replaceable>_<replaceable>VARIABLE</replaceable></varname> and <varname><replaceable>OPT</replaceable>_<replaceable>VARIABLE</replaceable>_OFF</varname> For any of these variables: ALL_TARGET BROKEN CATEGORIES CFLAGS CONFIGURE_ENV CONFLICTS CONFLICTS_BUILD CONFLICTS_INSTALL CPPFLAGS CXXFLAGS DESKTOP_ENTRIES DISTFILES EXTRA_PATCHES EXTRACT_ONLY GH_ACCOUNT GH_PROJECT GH_TAGNAME GH_TUPLE IGNORE INFO INSTALL_TARGET LDFLAGS LIBS MAKE_ARGS MAKE_ENV PATCHFILES PATCH_SITES PLIST_DIRS PLIST_DIRSTRY PLIST_FILES PLIST_SUB PORTDOCS PORTEXAMPLES SUB_FILES SUB_LIST TEST_TARGET USES When option OPT is selected, the value of OPT_ABOVEVARIABLE, if defined, is appended to ABOVEVARIABLE. OPT_ABOVEVARIABLE_OFF works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 OPT1_USES= gmake OPT1_CFLAGS_OFF= -DTEST is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MOPT1} USES+= gmake .else CFLAGS+= -DTEST .endif Some variables are not in this list, in particular PKGNAMEPREFIX and PKGNAMESUFFIX. This is intentional. A port must not change its name when its option set changes. Some of these variables, at least ALL_TARGET and INSTALL_TARGET, have their default values set after the options are processed. With these lines in the Makefile: ALL_TARGET= all DOCS_ALL_TARGET= doc If the DOCS option is enabled, ALL_TARGET will have a final value of all doc; if the option is disabled, it would have a value of all. With only the options helper line in the Makefile: DOCS_ALL_TARGET= doc If the DOCS option is enabled, ALL_TARGET will have a final value of doc; if the option is disabled, it would have a value of all. Additional Build Targets, <buildtarget><replaceable>TARGET</replaceable>-<replaceable>OPT</replaceable>-on</buildtarget> and <buildtarget><replaceable>TARGET</replaceable>-<replaceable>OPT</replaceable>-on</buildtarget> These Makefile targets can accept optional extra build targets: pre-fetch do-fetch post-fetch pre-extract do-extract post-extract pre-patch do-patch post-patch pre-configure do-configure post-configure pre-build do-build post-build pre-install do-install post-install post-stage pre-package do-package post-package When option OPT is selected, the target TARGET-OPT-on, if defined, is executed after TARGET. TARGET-OPT-off works the same way, but when OPT is not selected. For example: OPTIONS_DEFINE= OPT1 post-patch-OPT1-on: @${REINPLACE_CMD} -e '/opt1/d' ${WRKSRC}/Makefile post-patch-OPT1-off: @${REINPLACE_CMD} -e '/opt1/s|/usr/bin/|${LOCALBASE}/bin/|' ${WRKSRC}/Makefile is equivalent to: OPTIONS_DEFINE= OPT1 .include <bsd.port.options.mk> post-patch: .if ${PORT_OPTIONS:MOPT1} @${REINPLACE_CMD} -e '/opt1/d' ${WRKSRC}/Makefile .else @${REINPLACE_CMD} -e '/opt1/s|/usr/bin/|${LOCALBASE}/bin/|' ${WRKSRC}/Makefile .endif Specifying the Working Directory Each port is extracted into a working directory, which must be writable. The ports system defaults to having DISTFILES unpack in to a directory called ${DISTNAME}. In other words, if the Makefile has: PORTNAME= foo PORTVERSION= 1.0 then the port's distribution files contain a top-level directory, foo-1.0, and the rest of the files are located under that directory. A number of variables can be overridden if that is not the case. <varname>WRKSRC</varname> The variable lists the name of the directory that is created when the application's distfiles are extracted. If our previous example extracted into a directory called foo (and not foo-1.0) write: WRKSRC= ${WRKDIR}/foo or possibly WRKSRC= ${WRKDIR}/${PORTNAME} <varname>WRKSRC_SUBDIR</varname> If the source files needed for the port are in a subdirectory of the extracted distribution file, set WRKSRC_SUBDIR to that directory. WRKSRC_SUBDIR= src <varname>NO_WRKSUBDIR</varname> If the port does not extract in to a subdirectory at all, then set NO_WRKSUBDIR to indicate that. NO_WRKSUBDIR= yes Because WRKDIR is the only directory that is supposed to be writable during the build, and is used to store many files recording the status of the build, the port's extraction will be forced into a subdirectory. Conflict Handling There are three different variables to register a conflict between packages and ports: CONFLICTS, CONFLICTS_INSTALL and CONFLICTS_BUILD. The conflict variables automatically set the variable IGNORE, which is more fully documented in . When removing one of several conflicting ports, it is advisable to retain CONFLICTS in those other ports for a few months to cater for users who only update once in a while. <varname>CONFLICTS_INSTALL</varname> If the package cannot coexist with other packages (because of file conflicts, runtime incompatibilities, etc.), list the other package names in CONFLICTS_INSTALL. Use shell globs like * and ? here. Enumerate package names in there, not port names or origins. Please make sure that CONFLICTS_INSTALL does not match this port's package itself. Otherwise enforcing its installation with FORCE_PKG_REGISTER will no longer work. CONFLICTS_INSTALL check is done after the build stage and prior to the install stage. <varname>CONFLICTS_BUILD</varname> If the port cannot be built when other specific ports are already installed, list the other port names in CONFLICTS_BUILD. Use shell globs like * and ? here. Use package names, not port names or origins. CONFLICTS_BUILD check is done prior to the build stage. Build conflicts are not recorded in the resulting package. <varname>CONFLICTS</varname> If the port cannot be built if a certain port is already installed and the resulting package cannot coexist with the other package, list the other package name in CONFLICTS. use shell globs like * and ? here. Enumerate package names in there, not port names or origins. Please make sure that CONFLICTS does not match this port's package itself. Otherwise enforcing its installation with FORCE_PKG_REGISTER will no longer work. CONFLICTS check is done prior to the build stage and prior to the install stage. Installing Files + + The install phase is the most + important from the end-user's point of view as it is the one + adding files to his system. All the additional commands run + in the port's Makefile in all the + *-install targets should be + echo'ed. Do not silence them with + @ or .SILENT. + + <varname>INSTALL_<replaceable>*</replaceable></varname> Macros Use the macros provided in bsd.port.mk to ensure correct modes of files in the port's *-install targets. Set ownership directly in pkg-plist with the corresponding entries, such as @(owner,group,), @owner owner, and @group group. These operators work until overridden, or until the end of pkg-plist, so do not forget to reset them after they are no longer needed. The default ownership is root:wheel. See for more information. INSTALL_PROGRAM is a command to install binary executables. INSTALL_SCRIPT is a command to install executable scripts. INSTALL_LIB is a command to install shared libraries (but not static libraries). INSTALL_KLD is a command to install kernel loadable modules. Some architectures do not like having the modules stripped, so use this command instead of INSTALL_PROGRAM. INSTALL_DATA is a command to install sharable data, including static libraries. INSTALL_MAN is a command to install manpages and other documentation (it does not compress anything). These variables are set to the &man.install.1; command with the appropriate flags for each situation. Do not use INSTALL_LIB to install static libraries, because stripping them renders them useless. Use INSTALL_DATA instead. Stripping Binaries and Shared Libraries Installed binaries should be stripped. Do not strip binaries manually unless absolutely required. The INSTALL_PROGRAM macro installs and strips a binary at the same time. The INSTALL_LIB macro does the same thing to shared libraries. When a file must be stripped, but neither INSTALL_PROGRAM nor INSTALL_LIB macros are desirable, ${STRIP_CMD} strips the program or shared library. This is typically done within the post-install target. For example: post-install: ${STRIP_CMD} ${STAGEDIR}${PREFIX}/bin/xdl When multiple files need to be stripped: post-install: .for l in geometry media body track world ${STRIP_CMD} ${STAGEDIR}${PREFIX}/lib/lib${PORTNAME}-${l}.so.0 .endfor Use &man.file.1; on a file to determine if it has been stripped. Binaries are reported by &man.file.1; as stripped, or not stripped. Additionally, &man.strip.1; will detect programs that have already been stripped and exit cleanly. Installing a Whole Tree of Files Sometimes, a large number of files must be installed while preserving their hierarchical organization. For example, copying over a whole directory tree from WRKSRC to a target directory under PREFIX. Note that PREFIX, EXAMPLESDIR, DATADIR, and other path variables must always be prepended with STAGEDIR to respect staging (see ). Two macros exist for this situation. The advantage of using these macros instead of cp is that they guarantee proper file ownership and permissions on target files. The first macro, COPYTREE_BIN, will set all the installed files to be executable, thus being suitable for installing into PREFIX/bin. The second macro, COPYTREE_SHARE, does not set executable permissions on files, and is therefore suitable for installing files under PREFIX/share target. post-install: ${MKDIR} ${STAGEDIR}${EXAMPLESDIR} (cd ${WRKSRC}/examples && ${COPYTREE_SHARE} . ${STAGEDIR}${EXAMPLESDIR}) This example will install the contents of the examples directory in the vendor distfile to the proper examples location of the port. post-install: ${MKDIR} ${STAGEDIR}${DATADIR}/summer (cd ${WRKSRC}/temperatures && ${COPYTREE_SHARE} "June July August" ${STAGEDIR}${DATADIR}/summer) And this example will install the data of summer months to the summer subdirectory of a DATADIR. Additional find arguments can be passed via the third argument to COPYTREE_* macros. For example, to install all files from the first example except Makefiles, one can use these commands. post-install: ${MKDIR} ${STAGEDIR}${EXAMPLESDIR} (cd ${WRKSRC}/examples && \ ${COPYTREE_SHARE} . ${STAGEDIR}${EXAMPLESDIR} "! -name Makefile") These macros do not add the installed files to pkg-plist. They must be added manually. For optional documentation (PORTDOCS, see ) and examples (PORTEXAMPLES), the %%PORTDOCS%% or %%PORTEXAMPLES%% prefixes must be prepended in pkg-plist. Install Additional Documentation If the software has some documentation other than the standard man and info pages that is useful for the user, install it under DOCSDIR This can be done, like the previous item, in the post-install target. Create a new directory for the port. The directory name is DOCSDIR. This usually equals PORTNAME. However, if the user might want different versions of the port to be installed at the same time, the whole PKGNAME can be used. Since only the files listed in pkg-plist are installed, it is safe to always install documentation to STAGEDIR (see ). Hence .if blocks are only needed when the installed files are large enough to cause significant I/O overhead. post-install: ${MKDIR} ${STAGEDIR}${DOCSDIR} ${INSTALL_MAN} ${WRKSRC}/docs/xvdocs.ps ${STAGEDIR}${DOCSDIR} + + On the other hand, if there is a DOCS option in the port, + install the documentation in a + post-install-DOCS-on target. These + targets are described in . Here are some handy variables and how they are expanded by default when used in the Makefile: DATADIR gets expanded to PREFIX/share/PORTNAME. DATADIR_REL gets expanded to share/PORTNAME. DOCSDIR gets expanded to PREFIX/share/doc/PORTNAME. DOCSDIR_REL gets expanded to share/doc/PORTNAME. EXAMPLESDIR gets expanded to PREFIX/share/examples/PORTNAME. EXAMPLESDIR_REL gets expanded to share/examples/PORTNAME. The DOCS option only controls additional documentation installed in DOCSDIR. It does not apply to standard man pages and info pages. Things installed in DATADIR and EXAMPLESDIR are controlled by DATA and EXAMPLES options, respectively. These variables are exported to PLIST_SUB. Their values will appear there as pathnames relative to PREFIX if possible. That is, share/doc/PORTNAME will be substituted for %%DOCSDIR%% in the packing list by default, and so on. (See more on pkg-plist substitution here.) All conditionally installed documentation files and directories are included in pkg-plist with the %%PORTDOCS%% prefix, for example: %%PORTDOCS%%%%DOCSDIR%%/AUTHORS %%PORTDOCS%%%%DOCSDIR%%/CONTACT As an alternative to enumerating the documentation files in pkg-plist, a port can set the variable PORTDOCS to a list of file names and shell glob patterns to add to the final packing list. The names will be relative to DOCSDIR. Therefore, a port that utilizes PORTDOCS, and uses a non-default location for its documentation, must set DOCSDIR accordingly. If a directory is listed in PORTDOCS or matched by a glob pattern from this variable, the entire subtree of contained files and directories will be registered in the final packing list. If the DOCS option has been unset then files and directories listed in PORTDOCS would not be installed or added to port packing list. Installing the documentation at PORTDOCS as shown above remains up to the port itself. A typical example of utilizing PORTDOCS looks as follows: PORTDOCS= README.* ChangeLog docs/* The equivalents of PORTDOCS for files installed under DATADIR and EXAMPLESDIR are PORTDATA and PORTEXAMPLES, respectively. The contents of pkg-message are displayed upon installation. See the section on using pkg-message for details. pkg-message does not need to be added to pkg-plist. Subdirectories Under <varname>PREFIX</varname> Try to let the port put things in the right subdirectories of PREFIX. Some ports lump everything and put it in the subdirectory with the port's name, which is incorrect. Also, many ports put everything except binaries, header files and manual pages in a subdirectory of lib, which does not work well with the BSD paradigm. Many of the files must be moved to one of these directories: etc (setup/configuration files), libexec (executables started internally), sbin (executables for superusers/managers), info (documentation for info browser) or share (architecture independent files). See &man.hier.7; for details; the rules governing /usr pretty much apply to /usr/local too. The exception are ports dealing with USENET news. They may use PREFIX/news as a destination for their files. Index: head/en_US.ISO8859-1/books/porters-handbook/plist/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/porters-handbook/plist/chapter.xml (revision 48371) +++ head/en_US.ISO8859-1/books/porters-handbook/plist/chapter.xml (revision 48372) @@ -1,953 +1,953 @@ Advanced <filename>pkg-plist</filename> Practices Changing <filename>pkg-plist</filename> Based on Make Variables Some ports, particularly the p5- ports, need to change their pkg-plist depending on what options they are configured with (or version of perl, in the case of p5- ports). To make this easy, any instances in pkg-plist of %%OSREL%%, %%PERL_VER%%, and %%PERL_VERSION%% will be substituted appropriately. The value of %%OSREL%% is the numeric revision of the operating system (for example, 4.9). %%PERL_VERSION%% and %%PERL_VER%% is the full version number of perl (for example, 5.8.9). Several other %%VARS%% related to port's documentation files are described in the relevant section. To make other substitutions, set PLIST_SUB with a list of VAR=VALUE pairs and instances of %%VAR%% will be substituted with VALUE in pkg-plist. For instance, if a port installs many files in a version-specific subdirectory, use a placeholder for the version so that pkg-plist does not have to be regenerated every time the port is updated. For example: OCTAVE_VERSION= ${PORTREVISION} PLIST_SUB= OCTAVE_VERSION=${OCTAVE_VERSION} in the Makefile and use %%OCTAVE_VERSION%% wherever the version shows up in pkg-plist. When the port is upgraded, it will not be necessary to edit dozens (or in some cases, hundreds) of lines in pkg-plist. If files are installed conditionally on the options set in the port, the usual way of handling it is prefixing pkg-plist lines with a %%OPT%% for lines needed when the option is enabled, or %%NO_OPT%% when the option is disabled, and adding OPTIONS_SUB=yes to the Makefile. See for more information. For instance, if there are files that are only installed when the X11 option is enabled, and Makefile has: OPTIONS_DEFINE= X11 OPTIONS_SUB= yes In pkg-plist, put %%X11%% in front of the lines only being installed when the option is enabled, like this : %%X11%%bin/foo-gui This substitution will be done between the pre-install and do-install targets, by reading from PLIST and writing to TMPPLIST (default: WRKDIR/.PLIST.mktmp). So if the port builds PLIST on the fly, do so in or before pre-install. Also, if the port needs to edit the resulting file, do so in post-install to a file named TMPPLIST. Another way of modifying a port's packing list is based on setting the variables PLIST_FILES and PLIST_DIRS. The value of each variable is regarded as a list of pathnames to write to TMPPLIST along with PLIST contents. While names listed in PLIST_FILES and PLIST_DIRS are subject to %%VAR%% substitution as described above, it is better to use the ${VAR} directly. Except for that, names from PLIST_FILES will appear in the final packing list unchanged, while @dir will be prepended to names from PLIST_DIRS. To take effect, PLIST_FILES and PLIST_DIRS must be set before TMPPLIST is written, that is, in pre-install or earlier. From time to time, using OPTIONS_SUB is not enough. In those cases, adding a specific TAG to PLIST_SUB inside the Makefile with a special value of @comment, makes package tools to ignore the line. For instance, if some files are only installed when the X11 option is on and the architecture is i386: .include <bsd.port.pre.mk> .if ${PORT_OPTIONS:MX11} && ${ARCH} == "i386" PLIST_SUB+= X11I386="" .else PLIST_SUB+= X11I386="@comment " .endif Empty Directories Cleaning Up Empty Directories When being de-installed, a port has to remove empty directories it created. Most of these directories are removed automatically by &man.pkg.8;, but for directories created outside of ${PREFIX}, or empty directories, some more work needs to be done. This is usually accomplished by adding @dir lines for those directories. Subdirectories must be deleted before deleting parent directories. [...] @dir /var/games/oneko/saved-games @dir /var/games/oneko Creating Empty Directories Empty directories created during port installation need special attention. They must be present when the package is created. If they are not created by the port code, create them in the Makefile: post-stage: - @${MKDIR} ${STAGEDIR}${PREFIX}/some/directory + ${MKDIR} ${STAGEDIR}${PREFIX}/some/directory Add the directory to pkg-plist like any other. For example: @dir some/directory Configuration Files If the port installs configuration files to PREFIX/etc (or elsewhere) do not list them in pkg-plist. That will cause pkg delete to remove files that have been carefully edited by the user, and a re-installation will wipe them out. Instead, install sample files with a filename.sample extension. The @sample macro automates this, see for what it does exactly. For each sample file, add a line to pkg-plist: @sample etc/orbit.conf.sample If there is a very good reason not to install a working configuration file by default, only list the sample filename in pkg-plist, without the @sample followed by a space part, and add a message pointing out that the user must copy and edit the file before the software will work. When a port installs its configuration in a subdirectory of ${PREFIX}/etc, use ETCDIR, which defaults to ${PREFIX}/etc/${PORTNAME}, it can be overridden in the ports Makefile if there is a convention for the port to use some other directory. The %%ETCDIR%% macro will be used in its stead in pkg-plist. The sample configuration files should always have the .sample suffix. If for some historical reason using the standard suffix is not possible, or if the sample files come from some other directory, use this construct: @sample etc/orbit.conf-dist etc/orbit.conf or @sample %%EXAMPLESDIR%%/orbit.conf etc/orbit.conf The format is @sample sample-file actual-config-file. Dynamic Versus Static Package List A static package list is a package list which is available in the Ports Collection either as pkg-plist (with or without variable substitution), or embedded into the Makefile via PLIST_FILES and PLIST_DIRS. Even if the contents are auto-generated by a tool or a target in the Makefile before the inclusion into the Ports Collection by a committer (for example, using make makeplist>), this is still considered a static list, since it is possible to examine it without having to download or compile the distfile. A dynamic package list is a package list which is generated at the time the port is compiled based upon the files and directories which are installed. It is not possible to examine it before the source code of the ported application is downloaded and compiled, or after running a make clean. While the use of dynamic package lists is not forbidden, maintainers should use static package lists wherever possible, as it enables users to &man.grep.1; through available ports to discover, for example, which port installs a certain file. Dynamic lists should be primarily used for complex ports where the package list changes drastically based upon optional features of the port (and thus maintaining a static package list is infeasible), or ports which change the package list based upon the version of dependent software used. For example, ports which generate docs with Javadoc. Automated Package List Creation First, make sure the port is almost complete, with only pkg-plist missing. Running make makeplist will show an example for pkg-plist. The output of makeplist must be double checked for correctness as it tries to automatically guess a few things, and can get it wrong. User configuration files should be installed as filename.sample, as it is described in . info/dir must not be listed and appropriate install-info lines must be added as noted in the info files section. Any libraries installed by the port must be listed as specified in the shared libraries section. Expanding Package List with Keywords All keywords can also take optional arguments in parentheses. The arguments are owner, group, and mode. This argument is used on the file or directory referenced. To change the owner, group, and mode of a configuration file, use: @sample(games,games,640) etc/config.sample The arguments are optional. If only the group and mode need to be changed, use: @sample(,games,660) etc/config.sample <literal>@desktop-file-utils</literal> Will run update-desktop-database -q after installation and deinstallation. <literal>@fc</literal> <replaceable>directory</replaceable> Add a @dir entry for the directory passed as an argument, and run fc-cache -fs on that directory after installation and deinstallation. <literal>@fcfontsdir</literal> <replaceable>directory</replaceable> Add a @dir entry for the directory passed as an argument, and run fc-cache -fs, mkfontscale and mkfontdir on that directory after installation and deinstallation. Additionally, on deinstallation, it removes the fonts.scale and fonts.dir cache files if they are empty. This keyword is equivalent to adding both @fc directory and @fontsdir directory. <literal>@fontsdir</literal> <replaceable>directory</replaceable> Add a @dir entry for the directory passed as an argument, and run mkfontscale and mkfontdir on that directory after installation and deinstallation. Additionally, on deinstallation, it removes the fonts.scale and fonts.dir cache files if they are empty. <literal>@glib-schemas</literal> Runs glib-compile-schemas on installation and deinstallation. <literal>@info</literal> <replaceable>file</replaceable> Add the file passed as argument to the plist, and updates the info document index on installation and deinstallation. Additionally, it removes the index if empty on deinstallation. This should never be used manually, but always through INFO. See for more information. <literal>@kld</literal> <replaceable>directory</replaceable> Runs kldxref on the directory on installation and deinstallation. Additionally, on deinstallation, it will remove the directory if empty. <literal>@rmtry</literal> <replaceable>file</replaceable> Will remove the file on deinstallation, and not give an error if the file is not there. <literal>@sample</literal> <replaceable>file</replaceable> [<replaceable>file</replaceable>] This is used to handle installation of configuration files, through example files bundled with the package. The actual, non-sample, file is either the second filename, if present, or the first filename without the .sample extension. This does three things. First, add the first file passed as argument, the sample file, to the plist. Then, on installation, if the actual file is not found, copy the sample file to the actual file. And finally, on deinstallation, remove the actual file if it has not been modified. See for more information. <literal>@shared-mime-info</literal> <replaceable>directory</replaceable> Runs update-mime-database on the directory on installation and deinstallation. <literal>@shell</literal> <replaceable>file</replaceable> Add the file passed as argument to the plist. On installation, add the full path to file to /etc/shells, while making sure it is not added twice. On deinstallation, remove it from /etc/shells. <literal>@terminfo</literal> Do not use by itself. If the port installs *.terminfo files, add USES=terminfo to its Makefile. On installation and deinstallation, if tic is present, refresh ${PREFIX}/share/misc/terminfo.db from the *.terminfo files in ${PREFIX}/share/misc. Base Keywords There are a few keywords that are hardcoded, and documented in &man.pkg-create.8;. For the sake of completeness, they are also documented here. <literal>@</literal> [<replaceable>file</replaceable>] The empty keyword is a placeholder to use when the file's owner, group, or mode need to be changed. For example, to set the group of the file to games and add the setgid bit, add: @(,games,2755) sbin/daemon <literal>@preexec</literal> <replaceable>command</replaceable>, <literal>@postexec</literal> <replaceable>command</replaceable>, <literal>@preunexec</literal> <replaceable>command</replaceable>, <literal>@postunexec</literal> <replaceable>command</replaceable> Execute command as part of the package installation or deinstallation process. @preexec command Execute command as part of the pre-install scripts. @postexec command Execute command as part of the post-install scripts. @preunexec command Execute command as part of the pre-deinstall scripts. @postunexec command Execute command as part of the post-deinstall scripts. If command contains any of these sequences somewhere in it, they are expanded inline. For these examples, assume that @cwd is set to /usr/local and the last extracted file was bin/emacs. %F Expand to the last filename extracted (as specified). In the example case bin/emacs. %D Expand to the current directory prefix, as set with @cwd. In the example case /usr/local. %B Expand to the basename of the fully qualified filename, that is, the current directory prefix plus the last filespec, minus the trailing filename. In the example case, that would be /usr/local/bin. %f Expand to the filename part of the fully qualified name, or the converse of %B. In the example case, emacs. <literal>@mode</literal> <replaceable>mode</replaceable> Set default permission for all subsequently extracted files to mode. Format is the same as that used by &man.chmod.1;. Use without an arg to set back to default permissions (mode of the file while being packed). This must be a numeric mode, like 644, 4755, or 600. It cannnot be a relative mode like u+s. <literal>@owner</literal> <replaceable>user</replaceable> Set default ownership for all subsequent files to user. Use without an argument to set back to default ownership (root). <literal>@group</literal> <replaceable>group</replaceable> Set default group ownership for all subsequent files to group. Use without an arg to set back to default group ownership (wheel). <literal>@comment</literal> <replaceable>string</replaceable> This line is ignored when packing. <literal>@dir</literal> <replaceable>directory</replaceable> Declare directory name. By default, directories created under PREFIX by a package installation are automatically removed. Use this when an empty directory under PREFIX needs to be created, or when the directory needs to have non default owner, group, or mode. Directories outside of PREFIX need to be registered. For example, /var/db/${PORTNAME} needs to have a @dir entry whereas ${PREFIX}/share/${PORTNAME} does not if it contains files or uses the default owner, group, and mode. <literal>@exec</literal> <replaceable>command</replaceable>, <literal>@unexec</literal> <replaceable>command</replaceable> (Deprecated) Execute command as part of the installation or deinstallation process. Please use instead. <literal>@dirrm</literal> <replaceable>directory</replaceable> (Deprecated) Declare directory name to be deleted at deinstall time. By default, directories created under PREFIX by a package installation are deleted when the package is deinstalled. <literal>@dirrmtry</literal> <replaceable>directory</replaceable> (Deprecated) Declare directory name to be removed, as for @dirrm, but does not issue a warning if the directory cannot be removed. Creating New Keywords Package list files can be extended by keywords that are defined in the ${PORTSDIR}/Keywords directory. The settings for each keyword are stored in a UCL file named keyword.ucl. The file must contain at least one of these sections: attributes action pre-install post-install pre-deinstall post-deinstall pre-upgrade post-upgrade <literal>attributes</literal> Changes the owner, group, or mode used by the keyword. Contains an associative array where the possible keys are owner, group, and mode. The values are, respectively, a user name, a group name, and a file mode. For example: attributes: { owner: "games", group: "games", mode: 0555 } <literal>action</literal> Defines what happens to the keyword's parameter. Contains an array where the possible values are: setprefix Set the prefix for the next plist entries. dir Register a directory to be created on install and removed on deinstall. dirrm Register a directory to be deleted on deinstall. Deprecated. dirrmtry Register a directory to try and deleted on deinstall. Deprecated. file Register a file. setmode Set the mode for the next plist entries. setowner Set the owner for the next plist entries. setgroup Set the group for the next plist entries. comment Does not do anything, equivalent to not entering an action section. ignore_next Ignore the next entry in the plist. <literal>arguments</literal> If set to true, adds argument handling, splitting the whole line, %@, into numbered arguments, %1, %2, and so on. For example, for this line: @foo some.content other.content %1 and %2 will contain: some.content other.content It also affects how the action entry works. When there is more than one argument, the argument number must be specified. For example: actions: [file(1)] <literal>pre-install</literal>, <literal>post-install</literal>, <literal>pre-deinstall</literal>, <literal>post-deinstall</literal>, <literal>pre-upgrade</literal>, <literal>post-upgrade</literal> These keywords contains a &man.sh.1; script to be executed before or after installation, deinstallation, or upgrade of the package. In addition to the usual @exec %foo placeholders described in , there is a new one, %@, which represents the argument of the keyword. Custom Keyword Examples Example of a <literal>@dirrmtryecho</literal> Keyword This keyword does two things, it adds a @dirrmtry directory line to the packing list, and echoes the fact that the directory is removed when deinstalling the package. actions: [dirrmtry] post-deinstall: <<EOD echo "Directory %D/%@ removed." EOD Real Life Example, How <literal>@sample</literal> is Implemented This keyword does three things. It adds the first filename passed as an argument to @sample to the packing list, it adds to the post-install script instructions to copy the sample to the actual configuration file if it does not already exist, and it adds to the post-deinstall instructions to remove the configuration file if it has not been modified. actions: [file(1)] arguments: true post-install: <<EOD case "%1" in /*) sample_file="%1" ;; *) sample_file="%D/%1" ;; esac target_file="${sample_file%.sample}" set -- %@ if [ $# -eq 2 ]; then target_file=${2} fi case "${target_file}" in /*) target_file="${target_file}" ;; *) target_file="%D/${target_file}" ;; esac if ! [ -f "${target_file}" ]; then /bin/cp -p "${sample_file}" "${target_file}" && \ /bin/chmod u+w "${target_file}" fi EOD pre-deinstall: <<EOD case "%1" in /*) sample_file="%1" ;; *) sample_file="%D/%1" ;; esac target_file="${sample_file%.sample}" set -- %@ if [ $# -eq 2 ]; then set -- %@ target_file=${2} fi case "${target_file}" in /*) target_file="${target_file}" ;; *) target_file="%D/${target_file}" ;; esac if cmp -s "${target_file}" "${sample_file}"; then rm -f "${target_file}" else echo "You may need to manually remove ${target_file} if it is no longer needed." fi EOD Index: head/en_US.ISO8859-1/books/porters-handbook/special/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/porters-handbook/special/chapter.xml (revision 48371) +++ head/en_US.ISO8859-1/books/porters-handbook/special/chapter.xml (revision 48372) @@ -1,5539 +1,5539 @@ Special Considerations This section explains the most common things to consider when creating a port. Staging bsd.port.mk expects ports to work with a stage directory. This means that a port must not install files directly to the regular destination directories (that is, under PREFIX, for example) but instead into a separate directory from which the package is then built. In many cases, this does not require root privileges, making it possible to build packages as an unprivileged user. With staging, the port is built and installed into the stage directory, STAGEDIR. A package is created from the stage directory and then installed on the system. Automake tools refer to this concept as DESTDIR, but in &os;, DESTDIR has a different meaning (see ). No port really needs to be root. It can mostly be avoided by using USES=uidfix. If the port still runs commands like &man.chown.8;, &man.chgrp.1;, or forces owner or group with &man.install.1; then use USES=fakeroot to fake those calls. Some patching of the port's Makefiles will be needed. Meta ports, or ports that do not install files themselves but only depend on other ports, must avoid needlessly extracting the &man.mtree.8; to the stage directory. This is the basic directory layout of the package, and these empty directories will be seen as orphans. To prevent &man.mtree.8; extraction, add this line: NO_MTREE= yes Metaports should use USES=metaport. It sets up defaults for ports that do not fetch, build, or install anything. Staging is enabled by prepending STAGEDIR to paths used in the pre-install, do-install, and post-install targets (see the examples through the book). Typically, this includes PREFIX, ETCDIR, DATADIR, EXAMPLESDIR, MANPREFIX, DOCSDIR, and so on. Directories should be created as part of the post-install target. Avoid using absolute paths whenever possible. When creating a symlink, STAGEDIR is prepended to the target path only. For example: ${LN} -sf libfoo.so.42 ${STAGEDIR}${PREFIX}/lib/libfoo.so The source path ${PREFIX}/lib/libfoo.so.42 looks fine but could, in fact, be incorrect. Absolute paths can point to a wrong location, like when a remote file system has been mounted with NFS under a non-root mount point. Relative paths are less fragile, and often much shorter. Ports that install kernel modules must prepend STAGEDIR to their destination, by default /boot/modules. Bundled Libraries This section explains why bundled dependencies are considered bad and what to do about them. Why Bundled Libraries Are Bad Some software requires the porter to locate third-party libraries and add the required dependencies to the port. Other software bundles all necessary libraries into the distribution file. The second approach seems easier at first, but there are some serious drawbacks: This list is loosely based on the Fedora and Gentoo wikis, both licensed under the CC-BY-SA 3.0 license. Security If vulnerabilities are found in the upstream library and fixed there, they might not be fixed in the library bundled with the port. One reason could be that the author is not aware of the problem. This means that the porter must fix them, or upgrade to a non-vulnerable version, and send a patch to the author. This all takes time, which results in software being vulnerable longer than necessary. This in turn makes it harder to coordinate a fix without unnecessarily leaking information about the vulnerability. Bugs This problem is similar to the problem with security in the last paragraph, but generally less severe. Forking It is easier for the author to fork the upstream library once it is bundled. While convenient on first sight, it means that the code diverges from upstream making it harder to address security or other problems with the software. A reason for this is that patching becomes harder. Another problem of forking is that because code diverges from upstream, bugs get solved over and over again instead of just once at a central location. This defeats the idea of open source software in the first place. Symbol collision When a library is installed on the system, it might collide with the bundled version. This can cause immediate errors at compile or link time. It can also cause errors when running the program which might be harder to track down. The latter problem could be caused because the versions of the two libraries are incompatible. Licensing When bundling projects from different sources, license issues can arise more easily, especially when licenses are incompatible. Waste of resources Bundled libraries waste resources on several levels. It takes longer to build the actual application, especially if these libraries are already present on the system. At run-time, they can take up unnecessary memory when the system-wide library is already loaded by one program and the bundled library is loaded by another program. Waste of effort When a library needs patches for &os;, these patches have to be duplicated again in the bundled library. This wastes developer time because the patches might not apply cleanly. It can also be hard to notice that these patches are required in the first place. What to do About Bundled Libraries Whenever possible, use the unbundled version of the library by adding a LIB_DEPENDS to the port. If such a port does not exist yet, consider creating it. Only use bundled libraries if the upstream has a good track record on security and using unbundled versions leads to overly complex patches. In some very special cases, for example emulators, like Wine, a port has to bundle libraries, because they are in a different architecture, or they have been modified to fit the software's use. In that case, those libraries should not be exposed to other ports for linking. Add BUNDLE_LIBS=yes to the port's Makefile. This will tell &man.pkg.8; to not compute provided libraries. Always ask the &a.portmgr; before adding this to a port. Shared Libraries If the port installs one or more shared libraries, define a USE_LDCONFIG make variable, which will instruct a bsd.port.mk to run ${LDCONFIG} -m on the directory where the new library is installed (usually PREFIX/lib) during post-install target to register it into the shared library cache. This variable, when defined, will also facilitate addition of an appropriate @exec /sbin/ldconfig -m and @unexec /sbin/ldconfig -R pair into pkg-plist, so that a user who installed the package can start using the shared library immediately and de-installation will not cause the system to still believe the library is there. USE_LDCONFIG= yes The default directory can be overridden by setting USE_LDCONFIG to a list of directories into which shared libraries are to be installed. For example, if the port installs shared libraries into PREFIX/lib/foo and PREFIX/lib/bar use this in Makefile: USE_LDCONFIG= ${PREFIX}/lib/foo ${PREFIX}/lib/bar Please double-check, often this is not necessary at all or can be avoided through -rpath or setting LD_RUN_PATH during linking (see lang/moscow_ml for an example), or through a shell-wrapper which sets LD_LIBRARY_PATH before invoking the binary, like www/seamonkey does. When installing 32-bit libraries on 64-bit system, use USE_LDCONFIG32 instead. If the software uses autotools, and specifically libtool, add USES=libtool. When the major library version number increments in the update to the new port version, all other ports that link to the affected library must have their PORTREVISION incremented, to force recompilation with the new library version. Ports with Distribution Restrictions or Legal Concerns Licenses vary, and some of them place restrictions on how the application can be packaged, whether it can be sold for profit, and so on. It is the responsibility of a porter to read the licensing terms of the software and make sure that the &os; project will not be held accountable for violating them by redistributing the source or compiled binaries either via FTP/HTTP or CD-ROM. If in doubt, please contact the &a.ports;. In situations like this, the variables described in the next sections can be set. <varname>NO_PACKAGE</varname> This variable indicates that we may not generate a binary package of the application. For instance, the license may disallow binary redistribution, or it may prohibit distribution of packages created from patched sources. However, the port's DISTFILES may be freely mirrored on FTP/HTTP. They may also be distributed on a CD-ROM (or similar media) unless NO_CDROM is set as well. If the binary package is not generally useful, and the application must always be compiled from the source code, use NO_PACKAGE. For example, if the application has configuration information that is site specific hard coded into it at compile time, set NO_PACKAGE. Set NO_PACKAGE to a string describing the reason why the package cannot be generated. <varname>NO_CDROM</varname> This variable alone indicates that, although we are allowed to generate binary packages, we may put neither those packages nor the port's DISTFILES onto a CD-ROM (or similar media) for resale. However, the binary packages and the port's DISTFILES will still be available via FTP/HTTP. If this variable is set along with NO_PACKAGE, then only the port's DISTFILES will be available, and only via FTP/HTTP. Set NO_CDROM to a string describing the reason why the port cannot be redistributed on CD-ROM. For instance, use this if the port's license is for non-commercial use only. <varname>NOFETCHFILES</varname> Files defined in NOFETCHFILES are not fetchable from any of MASTER_SITES. An example of such a file is when the file is supplied on CD-ROM by the vendor. Tools which check for the availability of these files on MASTER_SITES have to ignore these files and not report about them. <varname>RESTRICTED</varname> Set this variable alone if the application's license permits neither mirroring the application's DISTFILES nor distributing the binary package in any way. Do not set NO_CDROM or NO_PACKAGE along with RESTRICTED, since the latter variable implies the former ones. Set RESTRICTED to a string describing the reason why the port cannot be redistributed. Typically, this indicates that the port contains proprietary software and that the user will need to manually download the DISTFILES, possibly after registering for the software or agreeing to accept the terms of an EULA. <varname>RESTRICTED_FILES</varname> When RESTRICTED or NO_CDROM is set, this variable defaults to ${DISTFILES} ${PATCHFILES}, otherwise it is empty. If only some of the distribution files are restricted, then set this variable to list them. <varname>LEGAL_TEXT</varname> If the port has legal concerns not addressed by the above variables, set LEGAL_TEXT to a string explaining the concern. For example, if special permission was obtained for &os; to redistribute the binary, this variable must indicate so. <filename>/usr/ports/LEGAL</filename> and <varname>LEGAL</varname> A port which sets any of the above variables must also be added to /usr/ports/LEGAL. The first column is a glob which matches the restricted distfiles. The second column is the port's origin. The third column is the output of make -VLEGAL. Examples The preferred way to state "the distfiles for this port must be fetched manually" is as follows: .if !exists(${DISTDIR}/${DISTNAME}${EXTRACT_SUFX}) IGNORE= may not be redistributed because of licensing reasons. Please visit some-website to accept their license and download ${DISTFILES} into ${DISTDIR} .endif This both informs the user, and sets the proper metadata on the user's machine for use by automated programs. Note that this stanza must be preceded by an inclusion of bsd.port.pre.mk. Building Mechanisms Building Ports in Parallel The &os; ports framework supports parallel building using multiple make sub-processes, which allows SMP systems to utilize all of their available CPU power, allowing port builds to be faster and more effective. This is achieved by passing -jX flag to &man.make.1; running on vendor code. This is the default build behavior of ports. Unfortunately, not all ports handle parallel building well and it may be required to explicitly disable this feature by adding the MAKE_JOBS_UNSAFE=yes variable. It is used when a port is known to be broken with -jX. <command>make</command>, <command>gmake</command>, <command>fmake</command>, and <command>imake</command> Several differing make implementations exist. Ported software often requires a particular implementation, like GNU make, known in &os; as gmake, or fmake, the legacy &os; make. If the port uses GNU make, add gmake to USES. If the legacy &os; make is needed, add fmake there. MAKE_CMD can be used to reference the specific command configured by the USES setting in the port's Makefile. In rare cases when more than one make implementation is listed in USES, the variables GMAKE (for the GNU version) or FMAKE (for the legacy &os; version) are available. Only use MAKE_CMD within the application Makefiles in WRKSRC to call the make implementation expected by the ported software. If the port is an X application that uses imake to create Makefiles from Imakefiles, set USES= imake.. See the USES=imake section of for more details. If the port's source Makefile has something other than all as the main build target, set ALL_TARGET accordingly. The same goes for install and INSTALL_TARGET. <command>configure</command> Script If the port uses the configure script to generate Makefile from Makefile.in, set GNU_CONFIGURE=yes. To give extra arguments to the configure script (the default argument is --prefix=${PREFIX} --infodir=${PREFIX}/${INFO_PATH} --mandir=${MANPREFIX}/man --build=${CONFIGURE_TARGET}), set those extra arguments in CONFIGURE_ARGS. Extra environment variables can be passed using CONFIGURE_ENV. Variables for Ports That Use <command>configure</command> Variable Means GNU_CONFIGURE The port uses configure script to prepare build. HAS_CONFIGURE Same as GNU_CONFIGURE, except default configure target is not added to CONFIGURE_ARGS. CONFIGURE_ARGS Additional arguments passed to configure script. CONFIGURE_ENV Additional environment variables to be set for configure script run. CONFIGURE_TARGET Override default configure target. Default value is ${MACHINE_ARCH}-portbld-freebsd${OSREL}.
Using <command>cmake</command> For ports that use CMake, define USES= cmake, or USES= cmake:outsource to build in a separate directory (see below). Variables for Ports That Use <command>cmake</command> Variable Means CMAKE_ARGS Port specific CMake flags to be passed to the cmake binary. CMAKE_BUILD_TYPE Type of build (CMake predefined build profiles). Default is Release, or Debug if WITH_DEBUG is set. CMAKE_ENV Environment variables to be set for the cmake binary. Default is ${CONFIGURE_ENV}. CMAKE_SOURCE_PATH Path to the source directory. Default is ${WRKSRC}.
Variables the Users Can Define for <command>cmake</command> Builds Variable Means CMAKE_VERBOSE Enable verbose build output. Default not set, unless BATCH or PACKAGE_BUILDING are set. CMAKE_NOCOLOR Disables color build output. Default not set, unless BATCH or PACKAGE_BUILDING are set.
CMake supports these build profiles: Debug, Release, RelWithDebInfo and MinSizeRel. Debug and Release profiles respect system *FLAGS, RelWithDebInfo and MinSizeRel will set CFLAGS to -O2 -g and -Os -DNDEBUG correspondingly. The lower-cased value of CMAKE_BUILD_TYPE is exported to PLIST_SUB and must be used if the port installs *.cmake depending on the build type (see deskutils/strigi for an example). Please note that some projects may define their own build profiles and/or force particular build type by setting CMAKE_BUILD_TYPE in CMakeLists.txt. To make a port for such a project respect CFLAGS and WITH_DEBUG, the CMAKE_BUILD_TYPE definitions must be removed from those files. Most CMake-based projects support an out-of-source method of building. The out-of-source build for a port can be requested by using the :outsource suffix. When enabled, CONFIGURE_WRKSRC, BUILD_WRKSRC and INSTALL_WRKSRC will be set to ${WRKDIR}/.build and this directory will be used to keep all files generated during configuration and build stages, leaving the source directory intact. <literal>USES= cmake</literal> Example This snippet demonstrates the use of CMake for a port. CMAKE_SOURCE_PATH is not usually required, but can be set when the sources are not located in the top directory, or if only a subset of the project is intended to be built by the port. USES= cmake:outsource CMAKE_SOURCE_PATH= ${WRKSRC}/subproject Using <command>scons</command> If the port uses SCons, define USE_SCONS=yes. Variables for Ports That Use <command>scons</command> Variable Means SCONS_ARGS Port specific SCons flags passed to the SCons environment. SCONS_BUILDENV Variables to be set in system environment. SCONS_ENV Variables to be set in SCons environment. SCONS_TARGET Last argument passed to SCons, similar to MAKE_TARGET.
To make third party SConstruct respect everything that is passed to SCons in SCONS_ENV (that is, most importantly, CC/CXX/CFLAGS/CXXFLAGS), patch SConstruct so build Environment is constructed like this: env = Environment(**ARGUMENTS) It may be then modified with env.Append and env.Replace. Using GNU Autotools Introduction The various GNU autotools provide an abstraction mechanism for building a piece of software over a wide variety of operating systems and machine architectures. Within the Ports Collection, an individual port can make use of these tools via a simple construct: USE_AUTOTOOLS= tool[:env] ... At the time of writing, tool can be one of autoconf, autoheader, automake, aclocal, libtoolize. It can also be one the older legacy of autoconf213, autoheader213, automake14, aclocal14. env is used to specify that the environmental variables are needed. It also adds a build dependency on the tool. The relevant tool is not ran as part of the run-autotools target. Multiple tools can be specified at once, either by including them all on a single line, or using the += Makefile construct. <command>libtool</command> and <command>libtoolize</command> Ports shipping with their own copy of libtool (search for a file named ltmain.sh) need to have USES=libtool. If a port has USE_AUTOTOOLS=libtoolize it probably also needs USES=libtool. See the USES=libtool section in for more details. <filename>libltdl.so</filename> Some ports make use of the libltdl.so library package, which is part of the libtool suite. Use of this library does not automatically necessitate the use of libtool itself. If the port needs libltdl.so, add a dependency on it: LIB_DEPENDS= libltdl.so:${PORTSDIR}/devel/libltdl <command>autoconf</command> and <command>autoheader</command> Some ports do not contain a configure script, but do contain an autoconf template in configure.ac. Use these assignments to let autoconf create the configure script, and also have autoheader create template headers for use by the configure script. USE_AUTOTOOLS= autoconf[:env] and USE_AUTOTOOLS= autoheader which also implies the use of autoconf. The additional optional variables AUTOCONF_ARGS and AUTOHEADER_ARGS can be overridden by the port Makefile if specifically requested. Most ports are unlikely to need this. See bsd.autotools.mk for further details. <command>automake</command> and <command>aclocal</command> Some packages only contain Makefile.am. These have to be converted into Makefile.in using automake, and the further processed by configure to generate an actual Makefile. Similarly, packages occasionally do not ship with an included aclocal.m4, again required to build the software. This can be achieved with aclocal, which scans configure.ac or configure.in. aclocal has a similar relationship to automake as autoheader does to autoconf, described in the previous section. aclocal implies the use of automake, thus we have: USE_AUTOTOOLS= automake[:env] and USE_AUTOTOOLS= aclocal As with autoconf and autoheader, both automake and aclocal have optional argument variables, AUTOMAKE_ARGS and ACLOCAL_ARGS respectively, which may be overridden by the port Makefile if required. Using GNU <literal>gettext</literal> Basic Usage If the port requires gettext, set USES= gettext, and the port will inherit a dependency on libintl.so from devel/gettext. Other values for gettext usage are listed in USES=gettext. A rather common case is a port using gettext and configure. Generally, GNU configure should be able to locate gettext automatically. USES= gettext GNU_CONFIGURE= yes If it ever fails to, hints at the location of gettext can be passed in CPPFLAGS and LDFLAGS as follows: USES= gettext CPPFLAGS+= -I${LOCALBASE}/include LDFLAGS+= -L${LOCALBASE}/lib GNU_CONFIGURE= yes Optional Usage Some software products allow for disabling NLS. For example, through passing to configure. In that case, the port must use gettext conditionally, depending on the status of the NLS option. For ports of low to medium complexity, use this idiom: GNU_CONFIGURE= yes OPTIONS_DEFINE= NLS OPTIONS_SUB= yes NLS_USES= gettext NLS_CONFIGURE_ENABLE= nls .include <bsd.port.mk> Or using the older way of using options: GNU_CONFIGURE= yes OPTIONS_DEFINE= NLS .include <bsd.port.options.mk> .if ${PORT_OPTIONS:MNLS} USES+= gettext PLIST_SUB+= NLS="" .else CONFIGURE_ARGS+= --disable-nls PLIST_SUB+= NLS="@comment " .endif .include <bsd.port.mk> The next item on the to-do list is to arrange so that the message catalog files are included in the packing list conditionally. The Makefile part of this task is already provided by the idiom. It is explained in the section on advanced pkg-plist practices. In a nutshell, each occurrence of %%NLS%% in pkg-plist will be replaced by @comment  if NLS is disabled, or by a null string if NLS is enabled. Consequently, the lines prefixed by %%NLS%% will become mere comments in the final packing list if NLS is off; otherwise the prefix will be just left out. Then insert %%NLS%% before each path to a message catalog file in pkg-plist. For example: %%NLS%%share/locale/fr/LC_MESSAGES/foobar.mo %%NLS%%share/locale/no/LC_MESSAGES/foobar.mo In high complexity cases, more advanced techniques may be needed, such as dynamic packing list generation. Handling Message Catalog Directories There is a point to note about installing message catalog files. The target directories for them, which reside under LOCALBASE/share/locale, must not be created and removed by a port. The most popular languages have their respective directories listed in PORTSDIR/Templates/BSD.local.dist. The directories for many other languages are governed by the devel/gettext port. Consult its pkg-plist and see whether the port is going to install a message catalog file for a unique language. Using <application>Perl</application> If MASTER_SITES is set to CPAN, the correct subdirectory is usually selected automatically. If the default subdirectory is wrong, CPAN/Module can be used to change it. MASTER_SITES can also be set to the old MASTER_SITE_PERL_CPAN, then the preferred value of MASTER_SITE_SUBDIR is the top-level hierarchy name. For example, the recommended value for p5-Module-Name is Module. The top-level hierarchy can be examined at cpan.org. This keeps the port working when the author of the module changes. The exception to this rule is when the relevant directory does not exist or the distfile does not exist in that directory. In such case, using author's id as MASTER_SITE_SUBDIR is allowed. The CPAN:AUTHOR macro can be used, which will be translated to the hashed author directory. For example, CPAN:AUTHOR will be converted to authors/id/A/AU/AUTHOR. When a port needs Perl support, it must set USES=perl5 with the optional USE_PERL5 described in the perl5 USES description. Read-Only Variables for Ports That Use <application>Perl</application> Read only variables Means PERL The full path of the Perl 5 interpreter, either in the system or installed from a port, but without the version number. Use this when the software needs the path to the Perl interpreter. To replace #!lines in scripts, use USES=shebangfix. PERL_VERSION The full version of Perl installed (for example, 5.8.9). PERL_LEVEL The installed Perl version as an integer of the form MNNNPP (for example, 500809). PERL_ARCH Where Perl stores architecture dependent libraries. Defaults to ${ARCH}-freebsd. PERL_PORT Name of the Perl port that is installed (for example, perl5). SITE_PERL Directory name where site specific Perl packages go. This value is added to PLIST_SUB.
Ports of Perl modules which do not have an official website must link to cpan.org in the WWW line of pkg-descr. The preferred URL form is http://search.cpan.org/dist/Module-Name/ (including the trailing slash). Do not use ${SITE_PERL} in dependency declarations. Doing so assumes that perl5.mk has been included, which is not always true. Ports depending on this port will have incorrect dependencies if this port's files move later in an upgrade. The right way to declare Perl module dependencies is shown in the example below. Perl Dependency Example p5-IO-Tee>=0.64:${PORTSDIR}/devel/p5-IO-Tee For Perl ports that install manual pages, the macro PERL5_MAN3 can be used inside pkg-plist. For example, lib/perl5/5.14/man/man3/AnyEvent::I3.3.gz can be replaced with %%PERL5_MAN3%%/AnyEvent::I3.3.gz There are no PERL5_MANx macros for the other sections (x in 1, 2 and 4 to 9) because those get installed in the regular directories. Using X11 X.Org Components The X11 implementation available in The Ports Collection is X.Org. If the application depends on X components, set USE_XORG to the list of required components. Available components, at the time of writing, are: bigreqsproto compositeproto damageproto dmx dmxproto dri2proto dri3proto evieproto fixesproto fontcacheproto fontenc fontsproto fontutil glproto ice inputproto kbproto libfs oldx pciaccess pixman presentproto printproto randrproto recordproto renderproto resourceproto scrnsaverproto sm trapproto videoproto x11 xau xaw xaw6 xaw7 xbitmaps xcb xcmiscproto xcomposite xcursor xdamage xdmcp xevie xext xextproto xf86bigfontproto xf86dgaproto xf86driproto xf86miscproto xf86rushproto xf86vidmodeproto xfixes xfont xfontcache xft xi xinerama xineramaproto xkbfile xkbui xmu xmuu xorg-macros xorg-server xp xpm xprintapputil xprintutil xproto xproxymngproto xrandr xrender xres xscrnsaver xshmfence xt xtrans xtrap xtst xv xvmc xxf86dga xxf86misc xxf86vm. Always up-to-date list can be found in /usr/ports/Mk/bsd.xorg.mk. The Mesa Project is an effort to provide free OpenGL implementation. To specify a dependency on various components of this project, use USE_GL. Valid options are: egl, gl, glesv2, glew, glu, glut, glw and linux. For backwards compatibility, the value of yes maps to glu. <varname>USE_XORG</varname> Example USE_XORG= xrender xft xkbfile xt xaw USE_GL= glu Variables for Ports That Use X USES= imake The port uses imake. XMKMF Set to the path of xmkmf if not in the PATH. Defaults to xmkmf -a.
Using X11-Related Variables # Use some X11 libraries USE_XORG= x11 xpm Ports That Require Motif If the port requires a Motif library, define USES= motif in the Makefile. Default Motif implementation is x11-toolkits/open-motif. Users can choose x11-toolkits/lesstif instead by setting WANT_LESSTIF in their make.conf. MOTIFLIB will be set by motif.mk to reference the appropriate Motif library. Please patch the source of the port to use ${MOTIFLIB} wherever the Motif library is referenced in the original Makefile or Imakefile. There are two common cases: If the port refers to the Motif library as -lXm in its Makefile or Imakefile, substitute ${MOTIFLIB} for it. If the port uses XmClientLibs in its Imakefile, change it to ${MOTIFLIB} ${XTOOLLIB} ${XLIB}. Note that MOTIFLIB (usually) expands to -L/usr/local/lib -lXm -lXp or /usr/local/lib/libXm.a, so there is no need to add -L or -l in front. X11 Fonts If the port installs fonts for the X Window System, put them in LOCALBASE/lib/X11/fonts/local. Getting a Fake <envar>DISPLAY</envar> with Xvfb Some applications require a working X11 display for compilation to succeed. This pose a problem for machines that operate headless. When this variable is used, the build infrastructure will start the virtual framebuffer X server. The working DISPLAY is then passed to the build. See USES=display for the possible arguments. USES= display Desktop Entries Desktop entries (a Freedesktop standard) provide a way to automatically adjust desktop features when a new program is installed, without requiring user intervention. For example, newly-installed programs automatically appear in the application menus of compatible desktop environments. Desktop entries originated in the GNOME desktop environment, but are now a standard and also work with KDE and Xfce. This bit of automation provides a real benefit to the user, and desktop entries are encouraged for applications which can be used in a desktop environment. Using Predefined <filename>.desktop</filename> Files Ports that include predefined *.desktop must include those files in pkg-plist and install them in the $LOCALBASE/share/applications directory. The INSTALL_DATA macro is useful for installing these files. Updating Desktop Database If a port has a MimeType entry in its portname.desktop, the desktop database must be updated after install and deinstall. To do this, define USES= desktop-file-utils. Creating Desktop Entries with <varname>DESKTOP_ENTRIES</varname> Desktop entries can be easily created for applications by using DESKTOP_ENTRIES. A file named name.desktop will be created, installed, and added to pkg-plist automatically. Syntax is: DESKTOP_ENTRIES= "NAME" "COMMENT" "ICON" "COMMAND" "CATEGORY" StartupNotify The list of possible categories is available on the Freedesktop website. StartupNotify indicates whether the application is compatible with startup notifications. These are typically a graphic indicator like a clock that appear at the mouse pointer, menu, or panel to give the user an indication when a program is starting. A program that is compatible with startup notifications clears the indicator after it has started. Programs that are not compatible with startup notifications would never clear the indicator (potentially confusing and infuriating the user), and must have StartupNotify set to false so the indicator is not shown at all. Example: DESKTOP_ENTRIES= "ToME" "Roguelike game based on JRR Tolkien's work" \ "${DATADIR}/xtra/graf/tome-128.png" \ "tome -v -g" "Application;Game;RolePlaying;" \ false Using GNOME Introduction This chapter explains the GNOME framework as used by ports. The framework can be loosely divided into the base components, GNOME desktop components, and a few special macros that simplify the work of port maintainers. While developing a port or changing one, please set DEVELOPER=yes in the environment or in /etc/make.conf. This causes the ports framework to enable additional checks. Using <varname>USE_GNOME</varname> Adding this variable to the port allows the use of the macros and components defined in bsd.gnome.mk. The code in bsd.gnome.mk adds the needed build-time, run-time or library dependencies or the handling of special files. GNOME applications under &os; use the USE_GNOME infrastructure. Include all the needed components as a space-separated list. The USE_GNOME components are divided into these virtual lists: basic components, GNOME 3 components and legacy components. If the port needs only GTK3 libraries, this is the shortest way to define it: USE_GNOME= gtk30 USE_GNOME components automatically add the dependencies they need. Please see for an exhaustive list of all USE_GNOME components and which other components they imply and their dependencies. Here is an example Makefile for a GNOME port that uses many of the techniques outlined in this document. Please use it as a guide for creating new ports. # $FreeBSD$ PORTNAME= regexxer PORTVERSION= 0.10 CATEGORIES= devel textproc gnome MASTER_SITES= GNOME MAINTAINER= kwm@FreeBSD.org COMMENT= Interactive tool for performing search and replace operations USES= gettext gmake pathfix pkgconfig tar:xz GNU_CONFIGURE= yes USE_GNOME= gnomeprefix intlhack gtksourceviewmm3 CPPFLAGS+= -I${LOCALBASE}/include LDFLAGS+= -L${LOCALBASE}/lib INSTALLS_ICONS= yes GLIB_SCHEMAS= org.regexxer.gschema.xml .include <bsd.port.mk> The USE_GNOME macro without any arguments does not add any dependencies to the port. USE_GNOME cannot be set after bsd.port.pre.mk. Variables This section explains which macros are available and how they are used. Like they are used in the above example. The has a more in-depth explanation. USE_GNOME has to be set for these macros to be of use. INSTALLS_ICONS GTK+ ports which install Freedesktop-style icons to ${LOCALBASE}/share/icons should use this macro to ensure that the icons are cached and will display correctly. The cache file is named icon-theme.cache. Do not include that file in pkg-plist. This macro handles that automatically. This macro is not needed for Qt, which use a internal method. GLIB_SCHEMAS List of all the glib schema files the port installs. The macro will add the files to the port plist and handle the registration of these files on install and deinstall. The glib schema files are written in XML and end with the gschema.xml extension. They are installed in the share/glib-2.0/schemas/ directory. These schema files contain all application config values with there default settings. The actual database used by the applications is built by glib-compile-schema, which is run by the GLIB_SCHEMAS macro. GLIB_SCHEMAS=foo.gschema.xml Do not add glib schemas to the pkg-plist. If they are listed in pkg-plist, they will not be registered and the applications might not work properly. GCONF_SCHEMAS List all the gconf schema files. The macro will add the schema files to the port plist and will handle their registration on install and deinstall. GConf is the XML-based database that virtually all GNOME applications use for storing their settings. These files are installed into the etc/gconf/schemas directory. This database is defined by installed schema files that are used to generate %gconf.xml key files. For each schema file installed by the port, there be an entry in the Makefile: GCONF_SCHEMAS=my_app.schemas my_app2.schemas my_app3.schemas Gconf schemas are listed in the GCONF_SCHEMAS macro rather than pkg-plist. If they are listed in pkg-plist, they will not be registered and the applications might not work properly. INSTALLS_OMF Open Source Metadata Framework (OMF) files are commonly used by GNOME 2 applications. These files contain the application help file information, and require special processing by ScrollKeeper/rarian. To properly register OMF files when installing GNOME applications from packages, make sure that omf files are listed in pkg-plist and that the port Makefile has INSTALLS_OMF defined: INSTALLS_OMF=yes When set, bsd.gnome.mk automatically scans pkg-plist and adds appropriate @exec and @unexec directives for each .omf to track in the OMF registration database. GNOME Components For further help with a GNOME port, look at some of the existing ports for examples. The &os; GNOME page has contact information if more help is needed. The components are divided into GNOME components that are currently in use and legacy components. If the component supports argument, they are listed between parenthesis in the description. The first is the default. "Both" is shown if the component defaults to adding to both build and run dependencies. GNOME Components Component Associated program Description atk accessibility/atk Accessibility toolkit (ATK) atkmm accessibility/atkmm c++ bindings for atk cairo graphics/cairo Vector graphics library with cross-device output support cairomm graphics/cairomm c++ bindings for cairo dconf devel/dconf Configuration database system (both, build, run) evolutiondataserver3 databases/evolution-data-server Data backends for the Evolution integrated mail/PIM suite gdkpixbuf2 graphics/gdk-pixbuf2 Graphics library for GTK+ glib20 devel/glib20 GNOME core library glib20 glibmm devel/glibmm c++ bindings for glib20 gnomecontrolcenter3 sysutils/gnome-control-center GNOME 3 Control Center gnomedesktop3 x11/gnome-desktop GNOME 3 desktop UI library gsound audio/gsound GObject library for playing system sounds (both, build, run) gtk-update-icon-cache graphics/gtk-update-icon-cache Gtk-update-icon-cache utility from the Gtk+ toolkit gtk20 x11-toolkits/gtk20 Gtk+ 2 toolkit gtk30 x11-toolkits/gtk30 Gtk+ 3 toolkit gtkmm20 x11-toolkits/gtkmm20 c++ bindings 2.0 for the gtk20 toolkit gtkmm24 x11-toolkits/gtkmm24 c++ bindings 2.4 for the gtk20 toolkit gtkmm30 x11-toolkits/gtkmm30 c++ bindings 3.0 for the gtk30 toolkit gtksourceview2 x11-toolkits/gtksourceview2 Widget that adds syntax highlighting to GtkTextView gtksourceview3 x11-toolkits/gtksourceview3 Text widget that adds syntax highlighting to the GtkTextView widget gtksourceviewmm3 x11-toolkits/gtksourceviewmm3 c++ bindings for the gtksourceview3 library gvfs devel/gvfs GNOME virtual file system intltool textproc/intltool Tool for internationalization (also see intlhack) introspection devel/gobject-introspection Basic introspection bindings and tools to generate introspection bindings. Most of the time :build is enough, :both/:run is only need for applications that use introspection bindings. (both, build, run) libgda5 databases/libgda5 Provides uniform access to different kinds of data sources libgda5-ui databases/libgda5-ui UI library from the libgda5 library libgdamm5 databases/libgdamm5 c++ bindings for the libgda5 library libgsf devel/libgsf Extensible I/O abstraction for dealing with structured file formats librsvg2 graphics/librsvg2 Library for parsing and rendering SVG vector-graphic files libsigc++20 devel/libsigc++20 Callback Framework for C++ libxml++26 textproc/libxml++26 c++ bindings for the libxml2 library libxml2 textproc/libxml2 XML parser library (both, build, run) libxslt textproc/libxslt XSLT C library (both, build, run) metacity x11-wm/metacity Window manager from GNOME nautilus3 x11-fm/nautilus GNOME file manager pango x11-toolkits/pango Open-source framework for the layout and rendering of i18n text pangomm x11-toolkits/pangomm c++ bindings for the pango library py3gobject3 devel/py3-gobject3 Python 3, GObject 3.0 bindings pygobject3 devel/py-gobject3 Python 2, GObject 3.0 bindings vte3 x11-toolkits/vte3 Terminal widget with improved accessibility and I18N support
GNOME Macro Components Component Description gnomeprefix Supply configure with some default locations. intlhack Same as intltool, but patches to make sure share/locale/ is used. Please only use when intltool alone is not enough. referencehack This macro is there to help splitting of the API or reference documentation into its own port.
GNOME Legacy Components Component Associated program Description atspi accessibility/at-spi Assistive Technology Service Provider Interface esound audio/esound Enlightenment sound package gal2 x11-toolkits/gal2 Collection of widgets taken from GNOME 2 gnumeric gconf2 devel/gconf2 Configuration database system for GNOME 2 gconfmm26 devel/gconfmm26 c++ bindings for gconf2 gdkpixbuf graphics/gdk-pixbuf Graphics library for GTK+ glib12 devel/glib12 glib 1.2 core library gnomedocutils textproc/gnome-doc-utils GNOME doc utils gnomemimedata misc/gnome-mime-data MIME and Application database for GNOME 2 gnomesharp20 x11-toolkits/gnome-sharp20 GNOME 2 interfaces for the .NET runtime gnomespeech accessibility/gnome-speech GNOME 2 text-to-speech API gnomevfs2 devel/gnome-vfs GNOME 2 Virtual File System gtk12 x11-toolkits/gtk12 Gtk+ 1.2 toolkit gtkhtml3 www/gtkhtml3 Lightweight HTML rendering/printing/editing engine gtkhtml4 www/gtkhtml4 Lightweight HTML rendering/printing/editing engine gtksharp20 x11-toolkits/gtk-sharp20 GTK+ and GNOME 2 interfaces for the .NET runtime gtksourceview x11-toolkits/gtksourceview Widget that adds syntax highlighting to GtkTextView libartgpl2 graphics/libart_lgpl Library for high-performance 2D graphics libbonobo devel/libbonobo Component and compound document system for GNOME 2 libbonoboui x11-toolkits/libbonoboui GUI frontend to the libbonobo component of GNOME 2 libgda4 databases/libgda4 Provides uniform access to different kinds of data sources libglade2 devel/libglade2 GNOME 2 glade library libgnome x11/libgnome Libraries for GNOME 2, a GNU desktop environment libgnomecanvas graphics/libgnomecanvas Graphics library for GNOME 2 libgnomekbd x11/libgnomekbd GNOME 2 keyboard shared library libgnomeprint print/libgnomeprint Gnome 2 print support library libgnomeprintui x11-toolkits/libgnomeprintui Gnome 2 print support library libgnomeui x11-toolkits/libgnomeui Libraries for the GNOME 2 GUI, a GNU desktop environment libgtkhtml www/libgtkhtml Lightweight HTML rendering/printing/editing engine libgtksourceviewmm x11-toolkits/libgtksourceviewmm c++ binding of GtkSourceView libidl devel/libIDL Library for creating trees of CORBA IDL file libsigc++12 devel/libsigc++12 Callback Framework for C++ libwnck x11-toolkits/libwnck Library used for writing pagers and taskslists libwnck3 x11-toolkits/libwnck3 Library used for writing pagers and taskslists orbit2 devel/ORBit2 High-performance CORBA ORB with support for the C language pygnome2 x11-toolkits/py-gnome2 Python bindings for GNOME 2 pygobject devel/py-gobject Python 2, GObject 2.0 bindings pygtk2 x11-toolkits/py-gtk2 Set of Python bindings for GTK+ pygtksourceview x11-toolkits/py-gtksourceview Python bindings for GtkSourceView 2 vte x11-toolkits/vte Terminal widget with improved accessibility and I18N support
Deprecated Components: Do Not Use Component Description HAVE_GNOME Deprecated, do not use. Was used to check if a component was installed. This was used for ports that did not have --enable/--disable switches for their configure script. But the building of parts of a port without a implicit request is discouraged. WANT_GNOME Deprecated, do not use. Was used by ports that needed USE_GNOME for optional dependencies, which where defined after bsd.port.pre.mk. Since USE_GNOME can be used after the inclusion of bsd.port.options.mk, there is little need for this macro any more. pangox-compat pangox-compat has been deprecated and split off from the pango package.
Using Qt Ports That Require Qt The Ports Collection provides support for Qt 4 and Qt 5 frameworks with USE_QTx, where x is 4 or 5. Set USE_QTx to the list of required Qt components (libraries, tools, plugins). The Qt 4 and Qt 5 frameworks are quite similar. The main difference is the set of supported components. The Qt framework exports a number of variables which can be used by ports, some of them listed below: Variables Provided to Ports That Use Qt QT_PREFIX Set to the path where Qt was installed (${LOCALBASE}). QMAKE Full path to qmake binary. LRELEASE Full path to lrelease utility. MOC Full path to moc. RCC Full path to rcc. UIC Full path to uic. QT_INCDIR Qt include directory. QT_LIBDIR Qt libraries path. QT_PLUGINDIR Qt plugins path.
When using the Qt framework, these settings are deployed: CONFIGURE_ARGS+= --with-qt-includes=${QT_INCDIR} \ --with-qt-libraries=${QT_LIBDIR} \ --with-extra-libs=${LOCALBASE}/lib \ --with-extra-includes=${LOCALBASE}/include CONFIGURE_ENV+= QTDIR="${QT_PREFIX}" QMAKE="${QMAKE}" \ MOC="${MOC}" RCC="${RCC}" UIC="${UIC}" \ QMAKESPEC="${QMAKESPEC}" PLIST_SUB+= QT_INCDIR=${QT_INCDIR_REL} \ QT_LIBDIR=${QT_LIBDIR_REL} \ QT_PLUGINDIR=${QT_PLUGINDIR_REL} Some configure scripts do not support the arguments above. To suppress modification ofCONFIGURE_ENV and CONFIGURE_ARGS, set QT_NONSTANDARD. Component Selection Individual Qt tool and library dependencies must be specified in USE_QTx. Every component can be suffixed with _build or _run, the suffix indicating whether the dependency on the component is at buildtime or runtime. If unsuffixed, the component will be depended on at both build- and runtime. Usually, library components are specified unsuffixed, tool components are mostly specified with the _build suffix and plugin components are specified with the _run suffix. The most commonly used components are listed below (all available components are listed in _USE_QT_ALL, _USE_QT4_ONLY, and _USE_QT5_ONLY in /usr/ports/Mk/bsd.qt.mk): Available Qt Library Components Name Description core core library (Qt 5 only) corelib core library (Qt 4 only) dbus Qt DBus library gui graphical user interface library network network library opengl Qt OpenGL library script script library sql SQL library testlib unit testing library webkit Qt WebKit library xml Qt XML library
To determine the libraries an application depends on, run ldd on the main executable after a successful compilation. Available Qt Tool Components Name Description qmake Makefile generator/build utility buildtools build tools (moc, rcc), needed for almost every Qt application (Qt 5 only) linguisttools localization tools: lrelease, lupdate (Qt 5 only) linguist localization tools: lrelease, lupdate (Qt 4 only) moc meta object compiler, needed for almost every Qt application at buildtime (Qt 4 only) rcc resource compiler, needed if the application comes with *.rc or *.qrc files (Qt 4 only) uic user interface compiler, needed if the application comes with *.ui files, in practice, every Qt application with a GUI (Qt 4 only)
Available Qt Plugin Components Name Description iconengines SVG icon engine plugin, needed if the application ships SVG icons (Qt 4 only) imageformats plugins for TGA, TIFF, and MNG image formats
Selecting Qt 4 Components In this example, the ported application uses the Qt 4 graphical user interface library, the Qt 4 core library, all of the Qt 4 code generation tools and Qt 4's Makefile generator. Since the gui library implies a dependency on the core library, corelib does not need to be specified. The Qt 4 code generation tools moc, uic and rcc, as well as the Makefile generator qmake are only needed at buildtime, thus they are specified with the _build suffix: USE_QT4= gui moc_build qmake_build rcc_build uic_build Using <command>qmake</command> If the application provides a qmake project file (*.pro), define USES= qmake along with USE_QTx. Note that USES= qmake already implies a build dependency on qmake, therefore the qmake component can be omitted from USE_QTx. Similar to CMake, qmake supports out-of-source builds, which can be enabled by specifying the outsource argument (see USES= qmake example). Variables for Ports That Use <command>qmake</command> Variable Means QMAKE_ARGS Port specific qmake flags to be passed to the qmake binary. QMAKE_ENV Environment variables to be set for the qmake binary. The default is ${CONFIGURE_ENV}. QMAKE_SOURCE_PATH Path to qmake project files (.pro). The default is ${WRKSRC} if an out-of-source build is requested, empty otherwise.
<literal>USES= qmake</literal> Example This snippet demonstrates the use of qmake for a Qt 4 port: USES= qmake:outsource USE_QT4= moc_build For a Qt 5 port: USES= qmake:outsource USE_QT5= buildtools_build Qt applications are often written to be cross-platform and often X11/Unix is not the platform they are developed on, which in turn leads to certain loose ends, like: Missing additional include paths. Many applications come with system tray icon support, but neglect to look for includes and/or libraries in the X11 directories. To add directories to qmake's include and library search paths via the command line, use: QMAKE_ARGS+= INCLUDEPATH+=${LOCALBASE}/include \ LIBS+=-L${LOCALBASE}/lib Bogus installation paths. Sometimes data such as icons or .desktop files are by default installed into directories which are not scanned by XDG-compatible applications. editors/texmaker is an example for this - look at patch-texmaker.pro in the files directory of that port for a template on how to remedy this directly in the qmake project file. Using KDE KDE 4 Variable Definitions If the application depends on KDE 4, set USE_KDE4 to the list of required components. _build and _run suffixes can be used to force components dependency type (for example, baseapps_run). If no suffix is set, a default dependency type will be used. To force both types, add the component twice with both suffixes (for example, automoc4_build automoc4_run). The most commonly used components are listed below (up-to-date components are documented at the top of /usr/ports/Mk/bsd.kde4.mk): Available KDE 4 Components Name Description kdehier Hierarchy of common KDE directories kdelibs KDE core libraries kdeprefix If set, port will be installed into ${KDE4_PREFIX} automoc4 Build tool to automatically generate moc files akonadi Storage server for KDE PIM data soprano Library for Resource Description Framework (RDF) strigi Strigi desktop search library libkcddb KDE CDDB (compact disc database) library libkcompactdisc KDE library for interfacing with audio CDs libkdeedu Libraries used by educational applications libkdcraw KDE LibRaw library libkexiv2 KDE Exiv2 library libkipi KDE Image Plugin Interface libkonq Konqueror core library libksane KDE SANE ("Scanner Access Now Easy") library pimlibs Personal information management libraries kate Advanced text editor framework marble Virtual globe and world atlas okular Universal document viewer korundum KDE Ruby bindings perlkde KDE Perl bindings pykde4 KDE Python bindings pykdeuic4 PyKDE user interface compiler smokekde KDE SMOKE libraries
KDE 4 ports are installed into KDE4_PREFIX. This is achieved by specifying the kdeprefix component, which overrides the default PREFIX. The ports, however, respect any PREFIX set via the MAKEFLAGS environment variable and/or make arguments. Currently KDE4_PREFIX is identical to the default PREFIX, ${LOCALBASE}. <varname>USE_KDE4</varname> Example This is a simple example for a KDE 4 port. USES= cmake:outsource instructs the port to utilize CMake, a configuration tool widely used by KDE 4 projects (see for detailed usage). USE_KDE4 brings dependency on KDE libraries and makes port using automoc4 at build stage. Required KDE components and other dependencies can be determined through configure log. USE_KDE4 does not imply USE_QT4. If a port requires some Qt 4 components, specify them in USE_QT4. USES= cmake:outsource USE_KDE4= kdelibs kdeprefix automoc4 USE_QT4= moc_build qmake_build rcc_build uic_build Using Java Variable Definitions If the port needs a Java™ Development Kit (JDK™) to either build, run or even extract the distfile, then define USE_JAVA. There are several JDKs in the ports collection, from various vendors, and in several versions. If the port must use one of these versions, define which one. The most current version, and &os; default is java/openjdk6. Variables Which May be Set by Ports That Use Java Variable Means USE_JAVA Define for the remaining variables to have any effect. JAVA_VERSION List of space-separated suitable Java versions for the port. An optional "+" allows specifying a range of versions (allowed values: 1.5[+] 1.6[+] 1.7[+]). JAVA_OS List of space-separated suitable JDK port operating systems for the port (allowed values: native linux). JAVA_VENDOR List of space-separated suitable JDK port vendors for the port (allowed values: freebsd bsdjava sun openjdk). JAVA_BUILD When set, add the selected JDK port to the build dependencies. JAVA_RUN When set, add the selected JDK port to the run dependencies. JAVA_EXTRACT When set, add the selected JDK port to the extract dependencies.
Below is the list of all settings a port will receive after setting USE_JAVA: Variables Provided to Ports That Use Java Variable Value JAVA_PORT The name of the JDK port (for example, java/openjdk6). JAVA_PORT_VERSION The full version of the JDK port (for example, 1.6.0). Only the first two digits of this version number are needed, use ${JAVA_PORT_VERSION:C/^([0-9])\.([0-9])(.*)$/\1.\2/}. JAVA_PORT_OS The operating system used by the JDK port (for example, 'native'). JAVA_PORT_VENDOR The vendor of the JDK port (for example, 'openjdk'). JAVA_PORT_OS_DESCRIPTION Description of the operating system used by the JDK port (for example, 'Native'). JAVA_PORT_VENDOR_DESCRIPTION Description of the vendor of the JDK port (for example, 'OpenJDK BSD Porting Team'). JAVA_HOME Path to the installation directory of the JDK (for example, '/usr/local/openjdk6'). JAVAC Path to the Java compiler to use (for example, '/usr/local/openjdk6/bin/javac'). JAR Path to the jar tool to use (for example, '/usr/local/openjdk6/bin/jar' or '/usr/local/bin/fastjar'). APPLETVIEWER Path to the appletviewer utility (for example, '/usr/local/openjdk6/bin/appletviewer'). JAVA Path to the java executable. Use this for executing Java programs (for example, '/usr/local/openjdk6/bin/java'). JAVADOC Path to the javadoc utility program. JAVAH Path to the javah program. JAVAP Path to the javap program. JAVA_KEYTOOL Path to the keytool utility program. JAVA_N2A Path to the native2ascii tool. JAVA_POLICYTOOL Path to the policytool program. JAVA_SERIALVER Path to the serialver utility program. RMIC Path to the RMI stub/skeleton generator, rmic. RMIREGISTRY Path to the RMI registry program, rmiregistry. RMID Path to the RMI daemon program rmid. JAVA_CLASSES Path to the archive that contains the JDK class files, ${JAVA_HOME}/jre/lib/rt.jar.
Use the java-debug make target to get information for debugging the port. It will display the value of many of the previously listed variables. Additionally, these constants are defined so all Java ports may be installed in a consistent way: Constants Defined for Ports That Use Java Constant Value JAVASHAREDIR The base directory for everything related to Java. Default: ${PREFIX}/share/java. JAVAJARDIR The directory where JAR files is installed. Default: ${JAVASHAREDIR}/classes. JAVALIBDIR The directory where JAR files installed by other ports are located. Default: ${LOCALBASE}/share/java/classes.
The related entries are defined in both PLIST_SUB (documented in ) and SUB_LIST. Building with Ant When the port is to be built using Apache Ant, it has to define USE_ANT. Ant is thus considered to be the sub-make command. When no do-build target is defined by the port, a default one will be set that runs Ant according to MAKE_ENV, MAKE_ARGS and ALL_TARGET. This is similar to the USES= gmake mechanism, which is documented in . Best Practices When porting a Java library, the port has to install the JAR file(s) in ${JAVAJARDIR}, and everything else under ${JAVASHAREDIR}/${PORTNAME} (except for the documentation, see below). To reduce the packing file size, reference the JAR file(s) directly in the Makefile. Use this statement (where myport.jar is the name of the JAR file installed as part of the port): PLIST_FILES+= ${JAVAJARDIR}/myport.jar When porting a Java application, the port usually installs everything under a single directory (including its JAR dependencies). The use of ${JAVASHAREDIR}/${PORTNAME} is strongly encouraged in this regard. It is up the porter to decide whether the port installs the additional JAR dependencies under this directory or uses the already installed ones (from ${JAVAJARDIR}). When porting a &java; application that requires an application server such as www/tomcat7 to run the service, it is quite common for a vendor to distribute a .war. A .war is a Web application ARchive and is extracted when called by the application. Avoid adding a .war to pkg-plist. It is not considered best practice. An application server will expand war archive, but not clean it up properly if the port is removed. A more desirable way of working with this file is to extract the archive, then install the files, and lastly add these files to pkg-plist. TOMCATDIR= ${LOCALBASE}/apache-tomcat-7.0 WEBAPPDIR= myapplication post-extract: @${MKDIR} ${WRKDIR}/${PORTDIRNAME} @${TAR} xf ${WRKDIR}/myapplication.war -C ${WRKDIR}/${PORTDIRNAME} do-install: cd ${WRKDIR} && \ ${INSTALL} -d -o ${WWWOWN} -g ${WWWGRP} ${TOMCATDIR}/webapps/${PORTDIRNAME} - @cd ${WRKDIR}/${PORTDIRNAME} && ${COPYTREE_SHARE} \* ${WEBAPPDIR}/${PORTDIRNAME} + cd ${WRKDIR}/${PORTDIRNAME} && ${COPYTREE_SHARE} \* ${WEBAPPDIR}/${PORTDIRNAME} Regardless of the type of port (library or application), the additional documentation is installed in the same location as for any other port. The JavaDoc tool is known to produce a different set of files depending on the version of the JDK that is used. For ports that do not enforce the use of a particular JDK, it is therefore a complex task to specify the packing list (pkg-plist). This is one reason why porters are strongly encouraged to use PORTDOCS. Moreover, even if the set of files that will be generated by javadoc can be predicted, the size of the resulting pkg-plist advocates for the use of PORTDOCS. The default value for DATADIR is ${PREFIX}/share/${PORTNAME}. It is a good idea to override DATADIR to ${JAVASHAREDIR}/${PORTNAME} for Java ports. Indeed, DATADIR is automatically added to PLIST_SUB (documented in ) so use %%DATADIR%% directly in pkg-plist. As for the choice of building Java ports from source or directly installing them from a binary distribution, there is no defined policy at the time of writing. However, people from the &os; Java Project encourage porters to have their ports built from source whenever it is a trivial task. All the features that have been presented in this section are implemented in bsd.java.mk. If the port needs more sophisticated Java support, please first have a look at the bsd.java.mk Subversion log as it usually takes some time to document the latest features. Then, if the needed support that is lacking would be beneficial to many other Java ports, feel free to discuss it on the &a.java;. Although there is a java category for PRs, it refers to the JDK porting effort from the &os; Java project. Therefore, submit the Java port in the ports category as for any other port, unless the issue is related to either a JDK implementation or bsd.java.mk. Similarly, there is a defined policy regarding the CATEGORIES of a Java port, which is detailed in . Web Applications, Apache and PHP Apache Variables for Ports That Use Apache USE_APACHE The port requires Apache. Possible values: yes (gets any version), 22, 24, 22-24, 22+, etc. The default APACHE version is 22. More details are available in ports/Mk/bsd.apache.mk and at wiki.freebsd.org/Apache/. APXS Full path to the apxs binary. Can be overridden in the port. HTTPD Full path to the httpd binary. Can be overridden in the port. APACHE_VERSION The version of present Apache installation (read-only variable). This variable is only available after inclusion of bsd.port.pre.mk. Possible values: 22, 24. APACHEMODDIR Directory for Apache modules. This variable is automatically expanded in pkg-plist. APACHEINCLUDEDIR Directory for Apache headers. This variable is automatically expanded in pkg-plist. APACHEETCDIR Directory for Apache configuration files. This variable is automatically expanded in pkg-plist.
Useful Variables for Porting Apache Modules MODULENAME Name of the module. Default value is PORTNAME. Example: mod_hello SHORTMODNAME Short name of the module. Automatically derived from MODULENAME, but can be overridden. Example: hello AP_FAST_BUILD Use apxs to compile and install the module. AP_GENPLIST Also automatically creates a pkg-plist. AP_INC Adds a directory to a header search path during compilation. AP_LIB Adds a directory to a library search path during compilation. AP_EXTRAS Additional flags to pass to apxs.
Web Applications Web applications must be installed into PREFIX/www/appname. This path is available both in Makefile and in pkg-plist as WWWDIR, and the path relative to PREFIX is available in Makefile as WWWDIR_REL. The user and group of web server process are available as WWWOWN and WWWGRP, in case the ownership of some files needs to be changed. The default values of both are www. Use WWWOWN?= myuser and WWWGRP?= mygroup if the port needs different values. This allows the user to override them easily. Do not depend on Apache unless the web app explicitly needs Apache. Respect that users may wish to run a web app on different web server than Apache. PHP Variables for Ports That Use PHP USE_PHP The port requires PHP. The value yes adds a dependency on PHP. The list of required PHP extensions can be specified instead. Example: pcre xml gettext DEFAULT_PHP_VER Selects which major version of PHP will be installed as a dependency when no PHP is installed yet. Default is 5. Possible values: 4, 5 IGNORE_WITH_PHP The port does not work with PHP of the given version. Possible values: 4, 5 USE_PHPIZE The port will be built as a PHP extension. USE_PHPEXT The port will be treated as a PHP extension, including installation and registration in the extension registry. USE_PHP_BUILD Set PHP as a build dependency. WANT_PHP_CLI Want the CLI (command line) version of PHP. WANT_PHP_CGI Want the CGI version of PHP. WANT_PHP_MOD Want the Apache module version of PHP. WANT_PHP_SCR Want the CLI or the CGI version of PHP. WANT_PHP_WEB Want the Apache module or the CGI version of PHP.
PEAR Modules Porting PEAR modules is a very simple process. Add USES=pear to the port's Makefile. The framework will install the relevant files in the right places and automatically generate the plist at install time. Example Makefile for PEAR Class PORTNAME= Date PORTVERSION= 1.4.3 CATEGORIES= devel www pear MAINTAINER= example@domain.com COMMENT= PEAR Date and Time Zone Classes USES= pear .include <bsd.port.mk> <application>Horde</application> Modules In the same way, porting Horde modules is a simple process. Add USES=horde to the port's Makefile. The framework will install the relevant files in the right places and automatically generate the plist at install time. The USE_HORDE_BUILD and USE_HORDE_RUN variables can be used to add buildtime and runtime dependencies on other Horde modules. See Mk/Uses/horde.mk for a complete list of available modules. Example Makefile for <application>Horde</application> Module PORTNAME= Horde_Core PORTVERSION= 2.14.0 CATEGORIES= devel www pear MAINTAINER= horde@FreeBSD.org COMMENT= Horde Core Framework libraries OPTIONS_DEFINE= KOLAB SOCKETS KOLAB_DESC= Enable Kolab server support SOCKETS_DESC= Depend on sockets PHP extension USES= horde USE_PHP= session USE_HORDE_BUILD= Horde_Role USE_HORDE_RUN= Horde_Role Horde_History Horde_Pack \ Horde_Text_Filter Horde_View KOLAB_USE= HORDE_RUN=Horde_Kolab_Server,Horde_Kolab_Session SOCKETS_USE= PHP=sockets .include <bsd.port.mk> Using Python The Ports Collection supports parallel installation of multiple Python versions. Ports must use a correct python interpreter, according to the user-settable PYTHON_VERSION. Most prominently, this means replacing the path to python executable in scripts with the value of PYTHON_CMD. Ports that install files under PYTHON_SITELIBDIR must use the pyXY- package name prefix, so their package name embeds the version of Python they are installed into. PKGNAMEPREFIX= ${PYTHON_PKGNAMEPREFIX} Most Useful Variables for Ports That Use Python USES=python The port needs Python. The minimal required version can be specified with values such as 2.7+. Version ranges can also be specified by separating two version numbers with a dash: USES=python:3.2-3.3 USE_PYTHON=distutils Use Python distutils for configuring, compiling, and installing. This is required when the port comes with setup.py. This overrides the do-build and do-install targets and may also override do-configure if GNU_CONFIGURE is not defined. USE_PYTHON=autoplist Create the packaging list automatically. This also requires USE_PYTHON=distutils to be set. USE_PYTHON=concurrent The port will use an unique prefix, typically PYTHON_PKGNAMEPREFIX for certain directories, such as EXAMPLESDIR and DOCSDIR and also will append a suffix, the python version from PYTHON_VER, to binaries and scripts to be installed. This allows ports to be installed for different Python versions at the same time, which otherwise would install conflicting files. PYTHON_PKGNAMEPREFIX Used as a PKGNAMEPREFIX to distinguish packages for different Python versions. Example: py27- PYTHON_SITELIBDIR Location of the site-packages tree, that contains installation path of Python (usually LOCALBASE). PYTHON_SITELIBDIR can be very useful when installing Python modules. PYTHONPREFIX_SITELIBDIR The PREFIX-clean variant of PYTHON_SITELIBDIR. Always use %%PYTHON_SITELIBDIR%% in pkg-plist when possible. The default value of %%PYTHON_SITELIBDIR%% is lib/python%%PYTHON_VERSION%%/site-packages PYTHON_CMD Python interpreter command line, including version number. PYNUMERIC Dependency line for numeric extension. PYNUMPY Dependency line for the new numeric extension, numpy. (PYNUMERIC is deprecated by upstream vendor). PYXML Dependency line for XML extension (not needed for Python 2.0 and higher as it is also in base distribution).
A complete list of available variables can be found in /usr/ports/Mk/Uses/python.mk. Makefile for a Simple <application>Python</application> Module PORTNAME= sample PORTVERSION= 1.2.3 CATEGORIES= devel MAINTAINER= john@doe.tld COMMENT= Python sample module USES= python USE_PYTHON= autoplist distutils .include <bsd.port.mk> Some Python applications claim to have DESTDIR support (which would be required for staging) but it is broken (Mailman up to 2.1.16, for instance). This can be worked around by recompiling the scripts. This can be done, for example, in the post-build target. Assuming the Python scripts are supposed to reside in PYTHONPREFIX_SITELIBDIR after installation, this solution can be applied: (cd ${STAGEDIR}${PREFIX} \ && ${PYTHON_CMD} ${PYTHON_LIBDIR}/compileall.py \ -d ${PREFIX} -f ${PYTHONPREFIX_SITELIBDIR:S;${PREFIX}/;;}) This recompiles the sources with a path relative to the stage directory, and prepends the value of PREFIX to the file name recorded in the byte-compiled output file by -d. -f is required to force recompilation, and the :S;${PREFIX}/;; strips prefixes from the value of PYTHONPREFIX_SITELIBDIR to make it relative to PREFIX. Using <application>Tcl/Tk</application> The Ports Collection supports parallel installation of multiple Tcl/Tk versions. Ports should try to support at least the default Tcl/Tk version and higher with USES=tcl. It is possible to specify the desired version of tcl by appending :xx, for example, USES=tcl:85. The Most Useful Read-Only Variables for Ports That Use <application>Tcl/Tk</application> TCL_VER chosen major.minor version of Tcl TCLSH full path of the Tcl interpreter TCL_LIBDIR path of the Tcl libraries TCL_INCLUDEDIR path of the Tcl C header files TK_VER chosen major.minor version of Tk WISH full path of the Tk interpreter TK_LIBDIR path of the Tk libraries TK_INCLUDEDIR path of the Tk C header files
See the USES=tcl and USES=tk of for a full description of those variables. A complete list of those variables is available in /usr/ports/Mk/Uses/tcl.mk. Using Emacs This section is yet to be written. Using Ruby Useful Variables for Ports That Use Ruby Variable Description USE_RUBY The port requires Ruby. USE_RUBY_EXTCONF The port uses extconf.rb to configure. USE_RUBY_SETUP The port uses setup.rb to configure. RUBY_SETUP Set to the alternative name of setup.rb. Common value is install.rb.
This table shows the selected variables available to port authors via the ports infrastructure. These variables are used to install files into their proper locations. Use them in pkg-plist as much as possible. Do not redefine these variables in the port. Selected Read-Only Variables for Ports That Use Ruby Variable Description Example value RUBY_PKGNAMEPREFIX Used as a PKGNAMEPREFIX to distinguish packages for different Ruby versions. ruby19- RUBY_VERSION Full version of Ruby in the form of x.y.z[.p]. 1.9.3.484 RUBY_SITELIBDIR Architecture independent libraries installation path. /usr/local/lib/ruby/site_ruby/1.9 RUBY_SITEARCHLIBDIR Architecture dependent libraries installation path. /usr/local/lib/ruby/site_ruby/1.9/amd64-freebsd10 RUBY_MODDOCDIR Module documentation installation path. /usr/local/share/doc/ruby19/patsy RUBY_MODEXAMPLESDIR Module examples installation path. /usr/local/share/examples/ruby19/patsy
A complete list of available variables can be found in /usr/ports/Mk/bsd.ruby.mk. Using SDL USE_SDL is used to autoconfigure the dependencies for ports which use an SDL based library like devel/sdl12 and graphics/sdl_image. These SDL libraries for version 1.2 are recognized: sdl: devel/sdl12 console: devel/sdl_console gfx: graphics/sdl_gfx image: graphics/sdl_image mixer: audio/sdl_mixer mm: devel/sdlmm net: net/sdl_net pango: x11-toolkits/sdl_pango sound: audio/sdl_sound ttf: graphics/sdl_ttf These SDL libraries for version 2.0 are recognized: sdl: devel/sdl20 gfx: graphics/sdl2_gfx image: graphics/sdl2_image mixer: audio/sdl2_mixer net: net/sdl2_net ttf: graphics/sdl2_ttf Therefore, if a port has a dependency on net/sdl_net and audio/sdl_mixer, the syntax will be: USE_SDL= net mixer The dependency devel/sdl12, which is required by net/sdl_net and audio/sdl_mixer, is automatically added as well. Using USE_SDL with entries for SDL 1.2, it will automatically: Add a dependency on sdl12-config to BUILD_DEPENDS Add the variable SDL_CONFIG to CONFIGURE_ENV Add the dependencies of the selected libraries to LIB_DEPENDS Using USE_SDL with entries for SDL 2.0, it will automatically: Add a dependency on sdl2-config to BUILD_DEPENDS Add the variable SDL2_CONFIG to CONFIGURE_ENV Add the dependencies of the selected libraries to LIB_DEPENDS Using <application>wxWidgets</application> This section describes the status of the wxWidgets libraries in the ports tree and its integration with the ports system. Introduction There are many versions of the wxWidgets libraries which conflict between them (install files under the same name). In the ports tree this problem has been solved by installing each version under a different name using version number suffixes. The obvious disadvantage of this is that each application has to be modified to find the expected version. Fortunately, most of the applications call the wx-config script to determine the necessary compiler and linker flags. The script is named differently for every available version. Majority of applications respect an environment variable, or accept a configure argument, to specify which wx-config script to call. Otherwise they have to be patched. Version Selection To make the port use a specific version of wxWidgets there are two variables available for defining (if only one is defined the other will be set to a default value): Variables to Select <application>wxWidgets</application> Versions Variable Description Default value USE_WX List of versions the port can use All available versions USE_WX_NOT List of versions the port cannot use None
The available wxWidgets versions and the corresponding ports in the tree are: Available <application>wxWidgets</application> Versions Version Port 2.4 x11-toolkits/wxgtk24 2.6 x11-toolkits/wxgtk26 2.8 x11-toolkits/wxgtk28
The versions starting from 2.5 also come in Unicode version and are installed by a slave port named like the normal one plus a -unicode suffix, but this can be handled with variables (see ). The variables in can be set to one or more of these combinations separated by spaces: <application>wxWidgets</application> Version Specifications Description Example Single version 2.4 Ascending range 2.4+ Descending range 2.6- Full range (must be ascending) 2.4-2.6
There are also some variables to select the preferred versions from the available ones. They can be set to a list of versions, the first ones will have higher priority. Variables to Select Preferred <application>wxWidgets</application> Versions Name Designed for WANT_WX_VER the port WITH_WX_VER the user
Component Selection There are other applications that, while not being wxWidgets libraries, are related to them. These applications can be specified in WX_COMPS. These components are available: Available <application>wxWidgets</application> Components Name Description Version restriction wx main library none contrib contributed libraries none python wxPython (Python bindings) 2.4-2.6 mozilla wxMozilla 2.4 svg wxSVG 2.6
The dependency type can be selected for each component by adding a suffix separated by a semicolon. If not present then a default type will be used (see ). These types are available: Available <application>wxWidgets</application> Dependency Types Name Description build Component is required for building, equivalent to BUILD_DEPENDS run Component is required for running, equivalent to RUN_DEPENDS lib Component is required for building and running, equivalent to LIB_DEPENDS
The default values for the components are detailed in this table: Default <application>wxWidgets</application> Dependency Types Component Dependency type wx lib contrib lib python run mozilla lib svg lib
Selecting <application>wxWidgets</application> Components This fragment corresponds to a port which uses wxWidgets version 2.4 and its contributed libraries. USE_WX= 2.4 WX_COMPS= wx contrib Unicode The wxWidgets library supports Unicode since version 2.5. In the ports tree both versions are available and can be selected with these variables: Variables to Select Unicode in <application>wxWidgets</application> Versions Variable Description Designed for WX_UNICODE The port works only with the Unicode version the port WANT_UNICODE The port works with both versions but prefers the Unicode one the port WITH_UNICODE The port will use the Unicode version the user WITHOUT_UNICODE The port will use the normal version if supported (when WX_UNICODE is not defined) the user
Do not use WX_UNICODE for ports that can use both Unicode and normal versions. If the port needs to use Unicode by default, define WANT_UNICODE instead. Detecting Installed Versions To detect an installed version, define WANT_WX. If it is not set to a specific version then the components will have a version suffix. HAVE_WX will be filled after detection. Detecting Installed <application>wxWidgets</application> Versions and Components This fragment can be used in a port that uses wxWidgets if it is installed, or an option is selected. WANT_WX= yes .include <bsd.port.pre.mk> .if defined(WITH_WX) || !empty(PORT_OPTIONS:MWX) || !empty(HAVE_WX:Mwx-2.4) USE_WX= 2.4 CONFIGURE_ARGS+= --enable-wx .endif This fragment can be used in a port that enables wxPython support if it is installed or if an option is selected, in addition to wxWidgets, both version 2.6. USE_WX= 2.6 WX_COMPS= wx WANT_WX= 2.6 .include <bsd.port.pre.mk> .if defined(WITH_WXPYTHON) || !empty(PORT_OPTIONS:MWXPYTHON) || !empty(HAVE_WX:Mpython) WX_COMPS+= python CONFIGURE_ARGS+= --enable-wxpython .endif Defined Variables These variables are available in the port (after defining one from ). Variables Defined for Ports That Use <application>wxWidgets</application> Name Description WX_CONFIG The path to the wxWidgets wx-config script (with different name) WXRC_CMD The path to the wxWidgets wxrc program (with different name) WX_VERSION The wxWidgets version that is going to be used (for example, 2.6) WX_UNICODE If not defined but Unicode is going to be used then it will be defined
Processing in <filename>bsd.port.pre.mk</filename> Define WX_PREMK to be able to use the variables right after including bsd.port.pre.mk. When defining WX_PREMK, then the version, dependencies, components and defined variables will not change if modifying the wxWidgets port variables after including bsd.port.pre.mk. Using <application>wxWidgets</application> Variables in Commands This fragment illustrates the use of WX_PREMK by running the wx-config script to obtain the full version string, assign it to a variable and pass it to the program. USE_WX= 2.4 WX_PREMK= yes .include <bsd.port.pre.mk> .if exists(${WX_CONFIG}) VER_STR!= ${WX_CONFIG} --release PLIST_SUB+= VERSION="${VER_STR}" .endif The wxWidgets variables can be safely used in commands when they are inside targets without the need of WX_PREMK. Additional <command>configure</command> Arguments Some GNU configure scripts cannot find wxWidgets with just the WX_CONFIG environment variable set, requiring additional arguments. WX_CONF_ARGS can be used for provide them. Legal Values for <varname>WX_CONF_ARGS</varname> Possible value Resulting argument absolute --with-wx-config=${WX_CONFIG} relative --with-wx=${LOCALBASE} --with-wx-config=${WX_CONFIG:T}
Using <application>Lua</application> This section describes the status of the Lua libraries in the ports tree and its integration with the ports system. Introduction There are many versions of the Lua libraries and corresponding interpreters, which conflict between them (install files under the same name). In the ports tree this problem has been solved by installing each version under a different name using version number suffixes. The obvious disadvantage of this is that each application has to be modified to find the expected version. But it can be solved by adding some additional flags to the compiler and linker. Version Selection A port using Lua only needs to have this line: USES= lua If a specific version of Lua is needed, instructions on how to select it are given in the USES=lua part of . Defined Variables These variables are available in the port. Variables Defined for Ports That Use <application>Lua</application> Name Description LUA_VER The Lua version that is going to be used (for example, 5.1) LUA_VER_STR The Lua version without the dots (for example, 51) LUA_PREFIX The prefix where Lua (and components) is installed LUA_SUBDIR The directory under ${PREFIX}/bin, ${PREFIX}/share and ${PREFIX}/lib where Lua is installed LUA_INCDIR The directory where Lua and tolua header files are installed LUA_LIBDIR The directory where Lua and tolua libraries are installed LUA_MODLIBDIR The directory where Lua module libraries (.so) are installed LUA_MODSHAREDIR The directory where Lua modules (.lua) are installed LUA_PKGNAMEPREFIX The package name prefix used by Lua modules LUA_CMD The path to the Lua interpreter LUAC_CMD The path to the Lua compiler
Using <command>iconv</command> After 2013-10-08 (254273), &os;  10-CURRENT and newer versions have a native iconv in the operating system. On earlier versions, converters/libiconv was used as a dependency. For software that needs iconv, define USES=iconv. &os; versions before 10-CURRENT on 2013-08-13 (254273) do not have a native iconv. On these earlier versions, a dependency on converters/libiconv will be added automatically. When a port defines USES=iconv, these variables will be available: Variable name Purpose Value before &os; 10-CURRENT 254273 (2013-08-13) Value after &os; 10-CURRENT 254273 (2013-08-13) ICONV_CMD Directory where the iconv binary resides ${LOCALBASE}/bin/iconv /usr/bin/iconv ICONV_LIB ld argument to link to libiconv (if needed) -liconv (empty) ICONV_PREFIX Directory where the iconv implementation resides (useful for configure scripts) ${LOCALBASE} /usr ICONV_CONFIGURE_ARG Preconstructed configure argument for configure scripts --with-libiconv-prefix=${LOCALBASE} (empty) ICONV_CONFIGURE_BASE Preconstructed configure argument for configure scripts --with-libiconv=${LOCALBASE} (empty) These two examples automatically populate the variables with the correct value for systems using converters/libiconv or the native iconv respectively: Simple <command>iconv</command> Usage USES= iconv LDFLAGS+= -L${LOCALBASE}/lib ${ICONV_LIB} <command>iconv</command> Usage with <command>configure</command> USES= iconv CONFIGURE_ARGS+=${ICONV_CONFIGURE_ARG} As shown above, ICONV_LIB is empty when a native iconv is present. This can be used to detect the native iconv and respond appropriately. Sometimes a program has an ld argument or search path hardcoded in a Makefile or configure script. This approach can be used to solve that problem: Fixing Hardcoded <literal>-liconv</literal> USES= iconv post-patch: @${REINPLACE_CMD} -e 's/-liconv/${ICONV_LIB}/' ${WRKSRC}/Makefile In some cases it is necessary to set alternate values or perform operations depending on whether there is a native iconv. bsd.port.pre.mk must be included before testing the value of ICONV_LIB: Checking for Native <command>iconv</command> Availability USES= iconv .include <bsd.port.pre.mk> post-patch: .if empty(ICONV_LIB) # native iconv detected @${REINPLACE_CMD} -e 's|iconv||' ${WRKSRC}/Config.sh .endif .include <bsd.port.post.mk> Using Xfce Ports that need Xfce libraries or applications set USES=xfce. Specific Xfce library and application dependencies are set with values assigned to USE_XFCE. They are defined in /usr/ports/Mk/Uses/xfce.mk. The possible values are: Values of <varname>USE_XFCE</varname> garcon sysutils/garcon libexo x11/libexo libgui x11-toolkits/libxfce4gui libmenu x11/libxfce4menu libutil x11/libxfce4util panel x11-wm/xfce4-panel thunar x11-fm/thunar xfconf x11/xfce4-conf <varname>USES=xfce</varname> Example USES= xfce USE_XFCE= libmenu Using Xfce's Own GTK3 Widgets In this example, the ported application uses the GTK3-specific widgets x11/libxfce4menu and x11/xfce4-conf. USES= xfce:gtk3 USE_XFCE= libmenu xfconf Xfce components included this way will automatically include any dependencies they need. It is no longer necessary to specify the entire list. If the port only needs x11-wm/xfce4-panel, use: USES= xfce USE_XFCE= panel There is no need to list the components x11-wm/xfce4-panel needs itself like this: USES= xfce USE_XFCE= libexo libmenu libutil panel However, Xfce components and non-Xfce dependencies of the port must be included explicitly. Do not count on an Xfce component to provide a sub-dependency other than itself for the main port. Using Mozilla Variables for Ports That Use Mozilla USE_GECKO Gecko backend the port can handle. Possible values: libxul (libxul.so), seamonkey (libgtkembedmoz.so, deprecated, must not be used any more). USE_FIREFOX The port requires Firefox as a runtime dependency. Possible values: yes (get default version), 40, 36, 35. Default dependency is on version 40. USE_FIREFOX_BUILD The port requires Firefox as a buildtime dependency. Possible values: see USE_FIREFOX. This automatically sets USE_FIREFOX and assigns the same value. USE_SEAMONKEY The port requires SeaMonkey as a runtime dependency. Possible values: yes (get default version), 20, 11 (deprecated, must not be used any more). Default dependency is on version 20. USE_SEAMONKEY_BUILD The port requires SeaMonkey as a buildtime dependency. Possible values: see USE_SEAMONKEY. This automatically sets USE_SEAMONKEY and assigns the same value. USE_THUNDERBIRD The port requires Thunderbird as a runtime dependency. Possible values: yes (get default version), 31, 30 (deprecated, must not be used any more). Default dependency is on version 31. USE_THUNDERBIRD_BUILD The port requires Thunderbird as a buildtime dependency. Possible values: see USE_THUNDERBIRD. This automatically sets USE_THUNDERBIRD and assigns the same value.
A complete list of available variables can be found in /usr/ports/Mk/bsd.gecko.mk. Using Databases Variables for Ports Using Databases Variable Means USE_BDB If variable is set to yes, add dependency on databases/db5 port. The variable may also be set to values: 48, 5 or 6. It is possible to declare a range of acceptable values, USE_BDB=48+ will find the highest installed version, and fall back to 4.8 if nothing else is installed. WANT_BDB_VER will always build this port with a specific version of the Berkely DB. INVALID_BDB_VER can be used to specify a version that cannot be used. USE_MYSQL If the variable is set to yes, add a dependency on the databases/mysql56-client port. An associated variable, WANT_MYSQL_VER, may be set to values such as 51, 55, or 60. Additionally to specify use of Percona, use 56p, or for MariaDB, use 53m, 55m or 100m. USE_PGSQL Retired in favor of USES=pgsql. USE_SQLITE If set to yes, add dependency on databases/sqlite3 port. The variable may also be set to 3 or 2, to add a dependency on 3.x or 2.x, respectively.
More details are available in bsd.database.mk. Starting and Stopping Services (<literal>rc</literal> Scripts) rc.d scripts are used to start services on system startup, and to give administrators a standard way of stopping, starting and restarting the service. Ports integrate into the system rc.d framework. Details on its usage can be found in the rc.d Handbook chapter. Detailed explanation of the available commands is provided in &man.rc.8; and &man.rc.subr.8;. Finally, there is an article on practical aspects of rc.d scripting. With a mythical port called doorman, which needs to start a doormand daemon. Add the following to the Makefile: USE_RC_SUBR= doormand Multiple scripts may be listed and will be installed. Scripts must be placed in the files subdirectory and a .in suffix must be added to their filename. Standard SUB_LIST expansions will be ran against this file. Use of the %%PREFIX%% and %%LOCALBASE%% expansions is strongly encouraged as well. More on SUB_LIST in the relevant section. As of &os; 6.1-RELEASE, local rc.d scripts (including those installed by ports) are included in the overall &man.rcorder.8; of the base system. An example simple rc.d script to start the doormand daemon: #!/bin/sh # $FreeBSD$ # # PROVIDE: doormand # REQUIRE: LOGIN # KEYWORD: shutdown # # Add these lines to /etc/rc.conf.local or /etc/rc.conf # to enable this service: # # doormand_enable (bool): Set to NO by default. # Set it to YES to enable doormand. # doormand_config (path): Set to %%PREFIX%%/etc/doormand/doormand.cf # by default. . /etc/rc.subr name=doormand rcvar=doormand_enable load_rc_config $name : ${doormand_enable:="NO"} : ${doormand_config="%%PREFIX%%/etc/doormand/doormand.cf"} command=%%PREFIX%%/sbin/${name} pidfile=/var/run/${name}.pid command_args="-p $pidfile -f $doormand_config" run_rc_command "$1" Unless there is a very good reason to start the service earlier, or it runs as a particular user (other than root), all ports scripts must use: REQUIRE: LOGIN If the startup script launches a daemon that must be shutdown, the following will trigger a stop of the service on system shutdown: KEYWORD: shutdown If the script is not starting a persistent service this is not necessary. For optional configuration elements the "=" style of default variable assignment is preferable to the ":=" style here, since the former sets a default value only if the variable is unset, and the latter sets one if the variable is unset or null. A user might very well include something like: doormand_flags="" in their rc.conf.local, and a variable substitution using ":=" would inappropriately override the user's intention. The _enable variable is not optional, and must use the ":" for the default. Pre-Commit Checklist Before contributing a port with an rc.d script, and more importantly, before committing one, please consult this checklist to be sure that it is ready. The devel/rclint port can check for most of these, but it is not a substitute for proper review. If this is a new file, does it have a .sh extension? If so, that must be changed to just file.in since rc.d files may not end with that extension. Does the file have a $FreeBSD$ tag? Do the name of the file (minus .in), the PROVIDE line, and $name all match? The file name matching PROVIDE makes debugging easier, especially for &man.rcorder.8; issues. Matching the file name and $name makes it easier to figure out which variables are relevant in rc.conf[.local]. It is also a policy for all new scripts, including those in the base system. Is the REQUIRE line set to LOGIN? This is mandatory for scripts that run as a non-root user. If it runs as root, is there a good reason for it to run prior to LOGIN? If not, it must run after so that local scrips can be loosely grouped to a point in &man.rcorder.8; after most everything in the base is already running. Does the script start a persistent service? If so, it must have KEYWORD: shutdown. Make sure there is no KEYWORD: &os; present. This has not been necessary nor desirable for years. It is also an indication that the new script was copy/pasted from an old script, so extra caution must be given to the review. If the script uses an interpreted language like perl, python, or ruby, make certain that command_interpreter is set appropriately, for example, for Perl, by adding PERL=${PERL} to SUB_LIST and using %%PERL%%. Otherwise, &prompt.root; service name stop will probably not work properly. See &man.service.8; for more information. Have all occurrences of /usr/local been replaced with %%PREFIX%%? Do the default variable assignments come after load_rc_config? Are there default assignments to empty strings? They should be removed, but double-check that the option is documented in the comments at the top of the file. Are things that are set in variables actually used in the script? Are options listed in the default name_flags things that are actually mandatory? If so, they must be in command_args. The option is a red flag (pardon the pun) here, since it is usually the option to “daemonize” the process, and therefore is actually mandatory. name_flags must never be included in command_args (and vice versa, although that error is less common). Does the script execute any code unconditionally? This is frowned on. Usually these things must be dealt with through a start_precmd. All boolean tests must use the checkyesno function. No hand-rolled tests for [Yy][Ee][Ss], etc. If there is a loop (for example, waiting for something to start) does it have a counter to terminate the loop? We do not want the boot to be stuck forever if there is an error. Does the script create files or directories that need specific permissions, for example, a pid that needs to be owned by the user that runs the process? Rather than the traditional &man.touch.1;/&man.chown.8;/&man.chmod.1; routine, consider using &man.install.1; with the proper command line arguments to do the whole procedure with one step. Adding Users and Groups Some ports require a particular user account to be present, usually for daemons that run as that user. For these ports, choose a unique UID from 50 to 999 and register it in ports/UIDs (for users) and ports/GIDs (for groups). The unique identification should be the same for users and groups. Please include a patch against these two files when requiring a new user or group to be created for the port. Then use USERS and GROUPS in Makefile, and the user will be automatically created when installing the port. USERS= pulse GROUPS= pulse pulse-access pulse-rt The current list of reserved UIDs and GIDs can be found in ports/UIDs and ports/GIDs. Ports That Rely on Kernel Sources Some ports (such as kernel loadable modules) need the kernel source files so that the port can compile. Here is the correct way to determine if the user has them installed: USES= kmod Apart from this check, the kmod feature takes care of most items that these ports need to take into account.