install:compile_cygwin

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
install:compile_cygwin [2008/08/25 12:41] – 92.252.45.110install:compile_cygwin [2010/12/03 09:30] (current) – old revision restored eros
Line 1: Line 1:
 ====== Compiling CWB from source under Windows ====== ====== Compiling CWB from source under Windows ======
  
-Compilation and installation of CWB version 2.2.b98+Compilation and installation of CWB version 3.0 
 + 
 + 
 + 
 + 
 +===== Status of the Cywgin port ===== 
 + 
 +The Cygwin port of the CWB is **experimental**.  While the source code compiles without error messages and basic CQP queries work well, it is currently not possible to process large corpora (e.g. in the 100M word range). 
 + 
 +The problems we have encountered may be due to limitations in the virtual memory management of Windows and the Cygwin emulation layer manage.  Apparently, a user process is limited to a 2 GB address space in Windows, and Cygwin seems to impose [[http://cygwin.com/cygwin-ug-net/setup-maxmem.html|further restrictions]].  Help from someone with more experience in Windows and Cygwin programming would be highly welcome, so that we can identify the true cause of the problems and try to find a workaround. 
 + 
 +In the long term, we hope to offer a native Windows port of the CWB.  Please join the [[http://devel.sslmit.unibo.it/mailman/listinfo/cwb|CWBdev mailing list]] if you would be interested to work on this port. 
  
  
Line 12: Line 24:
   * ''gcc''   * ''gcc''
   * ''libncurses-devel''   * ''libncurses-devel''
 +  * ''make''
 +  * ''perl''
  
 We recommend to install the simple text editor ''nano'' (from the //Editors// section) for editing configuration files, but you can also use your favourite Windows text editor. We recommend to install the simple text editor ''nano'' (from the //Editors// section) for editing configuration files, but you can also use your favourite Windows text editor.
 +
  
 ===== Getting the source code ===== ===== Getting the source code =====
  
-Get the source code from [[http://cwb.sourceforge.net/|here]], and unpack it:+Get the source code from [[http://sourceforge.net/projects/cwb/|here]], and unpack it:
  
 <code bash> <code bash>
-tar xf cwb-XXXXXX.tgz+tar xf cwb-XXXXXX.tgz (current version 2.2.b99-RC1)
 </code> </code>
  
Line 26: Line 41:
  
 <code bash> <code bash>
-cd cwb-3.0+cd cwb-XXXXXX (current version 2.2.b99-RC1)
 </code> </code>
 +
 +**Important note:** In principle, you can unpack the CWB source code anywhere you like, but //don't put it on a network drive// (we've encountered some weird errors there) and make sure that the directory path //does not contain blanks// (which will happen e.g. if you put the source code on your Desktop). The best solution is probably to keep the source code somewhere in the Cygwin directory tree, e.g. your Cygwin home directory. 
 +
  
 ===== Edit the makefile ===== ===== Edit the makefile =====
  
-First you need to set a few parameters in Makefile.inc using your favorite editor, we used nano:+First you need to set a few parameters in ''config.mk'' using your favourite text editor (but //not// Microsoft Word!).  If you have installed the ''nano'' package as recommended above, just type the following command:
  
 <code bash> <code bash>
-nano -w Makefile.inc+nano -w config.mk
 </code> </code>
  
-In the **platform** directive, insert //gnuwin32//+Otherwise, navigate to the ''cwb-XXXXXX'' directory in Windows Explorer and open the file ''config.mk'' with a text editor. 
 + 
 +In the **platform** directive, insert //cygwin//
  
 <file> <file>
-# currently available platforms: +#  
-#       unix            standard unix configuration (should work on most up-to-date Unix platforms) +# PLATFORM-SPECIFIC CONFIGURATION (OS and CPU type) 
-#       linux           i386-Linux +# 
-#       darwin          Mac OS X / Darwin [use one of the CPU-specific entries below!]+# Pre-defined platform configuration files: 
 +#       unix          standard Unix configuration [must set ENDIAN manually!] 
 +#       linux         i386-Linux (generic) 
 +#         linux-64       - configuration for 64-bit CPUs 
 +#         linux-opteron  - with optimimzation for AMD Opteron processor 
 +#       darwin        Mac OS X / Darwin [use one of the CPU-specific entries below]
 #         darwin-g4      - with optimization for PowerPC G4 processor #         darwin-g4      - with optimization for PowerPC G4 processor
 #         darwin-g5      - with optimization for PowerPC G5 processor #         darwin-g5      - with optimization for PowerPC G5 processor
-#         darwin-i386    - with configuration for i386-compatible processors +#         darwin-i386    - configuration for i386-compatible processors 
-#       solaris         SUN Solaris 2.x     (used at the IMS Stuttgart) +#         darwin-64      - 64-bit build on Intel Core2 and newer processors 
-#       gnuwin32        Windows95 / WinNT using Gnuwin32 (experimental)+#         darwin-core2   - optimised build for Core 2 CPU (requires Xcode 3.1) 
 +#       solaris       SUN Solaris 8 for SPARC CPU 
 +#       cygwin        Win32 build using Cygwin emulation layer (experimental)
 # #
-PLATFORM = gnuwin32+include $(TOP)/config/platform/cygwin
 </file> </file>
  
-The **site** directive has to be //cygwin//+The **site** directive also has to be changed to //cygwin//
  
 <file> <file>
-# SITE-SPECIFIC CONFIGURATION: load site-specific config file 
 # #
-# currently defined sites: +# SITE-SPECIFIC CONFIGURATION (installation path and other local settings)
-#       osx-fink        Mac OS X installation in Fink's /sw tree +
-#       osx-release        Mac OS X release build (static) +
-#       compling        Linux build on IKW's corpus server compling.cogsci.uos.de +
-#       compling-release   Linux release build on compling +
-#       linux-rpm       Binary RPM package on Linux (see rpm-linux.spec) +
-#       ims             IMS installation in /projekte/CWB/... tree +
-#       ims-public      IMS public installation in /usr/local/... tree +
-#       ims-release        Solaris release build (static) +
-#       ims-sunos-release  IMS release build for SunOS (no longer available) +
-#       es-sh           Darwin build on es-sh.unizh.ch (Sebastian Hoffmann) +
-#       es-sh-release   Darwin release build on es-sh.unizh.ch +
-#       osx-escorp      Configuration for corpus server at University of Zurich (H.-M. Lehmann) +
-#       cygwin           HBE Windows config (for experimentation with gnuwin32) +
-#       none            manual configuration +
-#       standard        standard installation in /usr/local and /corpora/c1 trees+
 # #
-# NB If there is no config file for your site, you must set SITE=none +# Pre-defined site configuration files: 
-# NB and specify values at least for the next three variables below +#       standard        standard configuration (installation in /usr/local tree) 
-# NB (more may be needed). +#       classic         "classic" configuration (CWB v2.2, uses /corpora/c1/registry) 
-SITE = cygwin+#       osx-fink        Mac OS X installation in Fink's /sw tree 
 +#       binary-release  Build binary package for release (static if possible, local install in build/ tree) 
 +#         osx-release     ... for Mac OS X 
 +#         linux-release   ... for i386 Linux 
 +#         solaris-release ... for SUN Solaris 2.x 
 +#         linux-rpm       ... build binary RPM package on Linux (together with rpm-linux.spec) 
 +#       cygwin          Win32 / Cygwin configuration (experimental) 
 +#        
 +include $(TOP)/config/site/cygwin
 </file> </file>
 +
 +
  
 ===== Compilation ===== ===== Compilation =====
  
-Then we compile the different components, one at a time. First we clean and check dependencies:+The easiest way to compile the CWB is to type 
 + 
 +<code bash> 
 +make all 
 +</code> 
 + 
 +at the command line, and go to fetch a cup of coffee (due to the overhead of the Cygwin emulation layer, compilation is much slower than on Unix systems). 
 + 
 +Since the Cygwin port is still experimental, it is probably a good idea to compile each component of the CWB separately. This will make it easier to recognise compilation errors and warnings. First, clean up any old files and check dependencies:
  
 <code bash> <code bash>
Line 90: Line 119:
 </code> </code>
  
-Then we compile the corpus library:+Then, compile the ''editline'' library used by CQP, which is included in the CWB source code distribution:
  
 <code bash> <code bash>
-make cl+make editline
 </code> </code>
  
-Then editline:+Now compile the corpus library:
  
 <code bash> <code bash>
-make editline+make cl
 </code> </code>
 +
  
 Then the utilities: Then the utilities:
Line 108: Line 138:
 </code> </code>
  
-Finally we compile cqp:+And finally CQP:
  
 <code bash> <code bash>
 make cqp make cqp
 </code> </code>
 +
 +You may also want to check that the manpages are up to date:
 +
 +<code bash>
 +make man
 +</code>
 +
 +
  
 ===== Installation ===== ===== Installation =====
Line 119: Line 157:
  
 <code bash> <code bash>
-sudo make install+make install
 </code> </code>
 +
 +If you have set up Cygwin with a separate administrator account, you may need to type ''sudo make install'' and enter the administrator password here.
  • install/compile_cygwin.1219660898.txt.gz
  • Last modified: 2008/08/25 12:41
  • by 92.252.45.110