Diff for /parser3/INSTALL between versions 1.31.6.5 and 1.58

version 1.31.6.5, 2003/07/22 12:01:54 version 1.58, 2024/11/04 04:37:43
Line 1 Line 1
 0.Assuming your perform your builds in $HOME directory   $Id$
   and suggested subdirectories.  
   
 1.What is the process to compile Parser3?  1. What is the process to compile Parser3?
   
   First, you must have LIBGC installed (conservative garbage collector for C++).    Just start ./buildall script and you would get $HOME/parser3install/bin/parser3 binary.
       libgc >= 6.1 [http://www.hpl.hp.com/personal/Hans_Boehm/gc/gc_source/gc6.1.tar.gz]  
       
     Installation sequence of this library to $HOME/gc:  
   
     cd libgc-x.x  
     ./configure --prefix=$HOME/gc --disable-threads --silent  
     make install  
   
     Then you can use these Parser3 configure options to specify the location and static/dynamic linkage with libgc:  
         --with-static-gc=$HOME/gc/lib  
         --with-shared-gc=$HOME/gc/lib  
   
   As most UNIX programs Parser3 follows the "standard":  
   
   mkdir $HOME/parser3project  
   cd $HOME/parser3project  
   gunzip -c parser3-x.x.tar.gz | tar xvf -  
   mv parser3-x.x parser3  
   cd parser3  
   ./configure --prefix=/usr/local  
   make  
     
   That renaming is needed for SQL drivers to be able to find   
   necessary include-file in Parser3 sources.  
   
   NOTE:    NOTE:
   On some system you should run 'gmake'.    In case you do not need XML support, use ./buildall --without-xml
   
   NOTE:    NOTE:
   By default your would get 'parser3' binary without support of these features:    In case you need apache parser module (DSO), use ./buildall --with-apache
         * xml(xdoc and xnode classes)  
   Read section (3) on getting xml-abled parser.  
   
   
   Read section (4) on configure options.  
   
 2.What is the process to install Parser3?  
   
   make install  
   
   Installation layout:  
     $prefix/bin/  
       parser3             -- CGI and command line Parser3 interpreter  
       auto.p.dist         -- configuration file sample,   
                              copy it to auto.p and adjust to your needs  
     $prefix/etc/  
       parser3.charsets/   -- charset definition files  
         koi8-r.cfg        -- cyrillic charset [KOI8-R encoding]  
         windows-1250.cfg  -- central europian charset [windows-1250 encoding]  
         windows-1251.cfg  -- cyrillic charset [windows-1251 encoding]  
         windows-1257.cfg  -- baltic charset [windows-1257 encoding]  
   
 3.What other libraries are needed to compile/install parser?   
   
 3a. In case you wish XML(xdoc and xnode classes) support, you need libraries  
       libxml2 >= 2.5.5 [ftp://xmlsoft.org/libxml2-2.5.5.tar.gz]  
       libxslt >= 1.0.28 [ftp://xmlsoft.org/libxslt-1.0.28.tar.gz]  
       glib >= 1.2.10 [ftp://ftp.gtk.org/pub/gtk/v1.2/glib-1.2.10.tar.gz]  
       gdome2 >= 0.7.4 [http://gdome2.cs.unibo.it/tarball/gdome2-0.7.4.tar.gz]  
     on your system, and specify path to them to Parser3 configure script.  
   
     Sources of libgdome library must be patched prior to compile. [library author bug confirmation pending]  
   
     Installation sequence of these libraries to $HOME/gnome:  
   
     libxml2:  
   
     cd libxml2-x.x.x  
     ./configure --prefix=$HOME/gnome --without-zlib --without-iconv --silent  
     make install  
   
     libxslt:  
   
     cd libxslt-x.x.x  
     ./configure --with-libxml-prefix=$HOME/gnome --prefix=$HOME/gnome --without-debug --without-debugger --silent  
     make install  
   
     glib:  
   
     cd glib-x.x.x    NOTE:
     ./configure --prefix=$HOME/gnome     If you have gc, prce, libxml and libxslt installed, you can use ./configure --with-xml
     make install  
   
     gdome2:  
   
     cd gdome2-x.x.x  
     patch -p0 < $HOME/parser3project/parser3/src/lib/patches/gdome2.patch  
     ./configure --with-glib-prefix=$HOME/gnome --disable-glibtest --with-libxml-prefix=$HOME/gnome --prefix=$HOME/gnome --silent  
     make install  
   
     And copy these additional headers into install includes directory  
   
     mkdir $HOME/gnome/include/libgdome/gdomecore  
     cp libgdome/gdomecore/gdome-xml-node.h $HOME/gnome/include/libgdome/gdomecore  
     cp libgdome/gdomecore/gdome-xml-document.h $HOME/gnome/include/libgdome/gdomecore  
       
   
     Parser3 configure options, choose whether you need static or dynamic linkage with xml libs:  
         --with-static-xml=$HOME/gnome  
         --with-shared-xml=$HOME/gnome  
   
     NOTE:  
     In case your glib as installed into some other prefix,  
     specify it's configuration file with  
         --with-glib-config=/path/to/your/glib-config  
     NOTE:  
     In case you are using SHARED xml libraries, you may want to link in direct path to them,  
     use this key to specify which linker key to use do do that:  
         --with-pathlink=LKEY    put dynamic libraries paths to binary  
                                 using linker key (-R, -rpath-link)  
   
 3b. In case you wish $mail:receive support you need libraries  
       glib >= 1.2.10 [ftp://ftp.gtk.org/pub/gtk/v1.2/glib-1.2.10.tar.gz]  
       gmime >= 1.0.1 [http://spruce.sourceforge.net/gmime/sources/gmime-1.0.5.tar.gz]  
     on your system, and specify path to them to Parser3 configure script.  
   
     You may have already installed glib at 3b step.  
   
     glib:  
   
     cd glib-x.x.x    NOTE:
     ./configure --prefix=$HOME/gnome     In case you later would experience problems with gcc runtime exception handling 
     make install    (most notable when reporting sql-related problems), pass this to buildall script:
     --with-sjlj-exceptions (HPUX is reported to have such problems)
   
     gmime:    NOTE:
     On some systems there are no "make" and you should run 'gmake', change buildall script accordingly then.
   
     cd gmime-x.x.x  2. What is the process to install Parser3?
     ./configure --prefix=$HOME/gnome  
     make install  
   
     Parser3 configure options, choose whether you need static or dynamic linkage with xml libs:    Copy files from $HOME/parser3install directory into your cgi-bin directory.
         --with-static-mailreceive=$HOME/gnome  
         --with-shared-mailreceive=$HOME/gnome  
   
     NOTE:    Then install Parser3 to handle documents, step-by-step instructions:
     In case your glib as installed into some other prefix,      http://www.parser.ru/en/docs/lang/install4apachecgi.htm" in English
     specify it's configuration file with      http://www.parser.ru/docs/lang/install4apachecgi.htm" in Russian
         --with-glib-config=/path/to/your/glib-config  
     NOTE:    Directory layout:
     In case you are using SHARED mime library, you may want to link in direct path to it,      bin/
     use this key to specify which linker key to use do do that:        parser3               -- CGI and command line Parser3 interpreter
         --with-pathlink=LKEY    put dynamic libraries paths to binary        auto.p.dist           -- configuration file sample,
                                 using linker key (-R, -rpath-link)                                 copy it to auto.p and adjust to your needs
       share/
         charsets/
           parser3.charsets/   -- charset definition files
             koi8-r.cfg        -- cyrillic charset [KOI8-R encoding]
             windows-1251.cfg  -- cyrillic charset [windows-1251 encoding]
             ...
   
 4.  3. I have heard about $mail:receive experimental support, how do I use it?
   
   In case you're not root on machine you're installing Parser,    Just start ./buildall --with-mailreceive.
   we recommend you to do:    If you have glib and gmime installed, you can run ./configure --with-mailreceive.
   ./configure --prefix=/your/home/parser3install  
   and afterwards copy parser3install/bin/parser3 to your CGI directory  
   or make a symlink [if your web server configured to follow symlinks].  
   
   You can disable any exec operations by setting  4. Safety features
   --disable-execs    You can disable any exec operations by setting --disable-execs option.
   option to configure.  
   file::exec, file::cgi and mail:send (unix version) methods would be disabled.    file::exec, file::cgi and mail:send (unix version) methods would be disabled.
   
   You can enable reading and executing files, not belonging to group+user     You can enable reading and executing files, not belonging to group+user 
   other then effective by setting    other than effective by setting --disable-safe-mode option.
   --disable-safe-mode  
   option to configure.  
     
   You can disable user-configured sendmail commands by forcing it, setting    You can disable user-configured sendmail commands by forcing it, setting
   "--with=sendmail=COMMAND"    "--with=sendmail=COMMAND" option.
   option to configure.  
   
   In case you are experiencing problem with gcc runtime exception handling,  5. Since Parser 3.4.0 several optimizations were implemented.
   try to configure with    If you experience problems you can try to disable them by commenting corresponding
   --with-sjlj-exceptions    defines and recompiling parser3.
   option.    
     src/include/pa_opcode.h
   By default libstdc++ is linked statically, you can re-enable dynamic libstdc++ linkage by specifying      #define OPTIMIZE_BYTECODE_GET_CLASS               -- $a:
   --with-dynamic-stdcpp      #define OPTIMIZE_BYTECODE_GET_ELEMENT             -- $a ^a
   option to configure.        #define OPTIMIZE_BYTECODE_GET_OBJECT_ELEMENT      -- $a.b ^a.b
       #define OPTIMIZE_BYTECODE_GET_OBJECT_VAR_ELEMENT  -- $a.$b ^a.$b
       #define OPTIMIZE_BYTECODE_GET_SELF_ELEMENT        -- $self.a ^self.a
   Some root config options can be specified at configure time, and would override autodetects.      #define OPTIMIZE_BYTECODE_CONSTRUCT               -- $a(expr),  $a[value]
                                                            $.a(expr), $.a[value]
   List of enabled charsets:                                                           $self.a(expr), $self.a[value]
   --with-charsets=CHARSET[,CHARSET,...]  Enables charsets in root config (windows-1251,      #define OPTIMIZE_BYTECODE_CUT_REM_OPERATOR        -- cut rem operator with any number of params during compilation
                           windows-1250 windows-1257 koi8-r; Default is windows-1251)      #define OPTIMIZE_BYTECODE_STRING_POOL             -- simplifying string's bytecode into expression
   SQL clients:      #define OPTIMIZE_BYTECODE_GET_SELF_ELEMENT        -- bytecode for $self.a ^self.a
   --with-mysql-client=mysqlclientlib?params      #define OPTIMIZE_BYTECODE_GET_ELEMENT__SPECIAL    -- bytecode for .CLASS, .CLASS_NAME
   --with-pgsql-client=pgsqlclientlib?params      #define FEATURE_GET_ELEMENT4CALL                  -- ^o.m[] calls get_element4call("m"), not get_element("m")
   --with-oracle-client=oracleclientlib?params  
       src/include/pa_string.h
        #define HASH_CODE_CACHING  -- calculated hash codes are cached and used for sequential hash lookups
 Alexandr Petrosian <paf@design.ru> (http://paf.design.ru)  
     src/include/pa_memory.h
       #define USE_DESTRUCTORS    -- destructors are used to decrease memory consumption during code
                                     processing and make ^memory:compact[] calls non-essential.
   
     src/types/pa_method.h
       #define OPTIMIZE_CALL      -- allows faster operators execution by eliminating method frame
                                     with local variables creation and extra write context switches.
       #define OPTIMIZE_RESULT    -- parser methods are marked when the $result variable was used.
                                     This allows write operations and context switching optimization.
   
     src/types/pa_wwrapper.h
       #define OPTIMIZE_SINGLE_STRING_WRITE  -- reuse original VString in single string assignments.
   
     src/lib/cord/include/cord.h
       #define CORD_CAT_OPTIMIZATION -- CORD library never modifies source concatenations. But in parser
                                        write operations it is safe to modify them and save some memory.
       #define CORD_CHARS_CACHE      -- language cords with same language and length are cached and reused.
   
 $Id$    src/include/pa_string.h
       #define STRING_LENGTH_CACHING -- cache String::Body.length() for char* strings
   
     src/include/pa_hash.h
       #define HASH_ORDER            -- hash keys are iterated in the order of insertion

Removed from v.1.31.6.5  
changed lines
  Added in v.1.58


E-mail: