toolchain.en.yhtml2

include homepage.en.yhtml2

page "The YML Toolchain" {
    h1 > The Toolchain

    p   >>
        There are to possibilities to handle YML files: the YML compiler and the YML/YSLT
        processor.
        >>

    h2 id=compiler > The YML compiler

    p   >>
        The YML ¬http://en.wikipedia.org/wiki/Compiler compiler¬ is a small
        ¬http://www.python.org Python script¬. It provides the
        ¬http://en.wikipedia.org/wiki/Command-line_interface command line¬ front end «yml2c».
        As default, it compiles your ¬http://en.wikipedia.org/wiki/Scripting_language script¬
        and outputs to ¬http://en.wikipedia.org/wiki/Stdout#Standard_output_.28stdout.29 stdout¬,
        that usually is the ¬http://en.wikipedia.org/wiki/Computer_terminal terminal¬.
        Your ¬http://en.wikipedia.org/wiki/Shell_(computing) shell¬ provides options to redirect
        the output into a ¬http://en.wikipedia.org/wiki/Pipeline_(Unix) pipe¬ or
        a ¬http://en.wikipedia.org/wiki/Computer_file file¬.
        >>

    p > So to use it, just enter «yml2c» followed by the filename of your YML script:

    Code | % yml2c myscript.yml2 > myscript.result

    p > The «yml2c» command has the following options:

    h3 > -h, --help

    p > Show a help message and exit.

    h3 > -C, --old-syntax

    p > Use the syntax of YML 2 version 1.x (compatibility mode).

    h3 > -D, --emit-linenumbers

    p > Emit line numbers into the resulting XML for debugging purposes.

    h3 > -E ENCODING, --encoding ENCODING

    p > encoding of input files (default: UTF-8)

    h3 > -I INCLUDE_PATH, --include=INCLUDE_PATH

    p { "Precede " code > ¬features#ymlpath YML_PATH¬
      " by a colon separated " code "INCLUDE_PATH"; " (semicolon separated on Windows) to search for include files." }

    h3 > -m, --omit-empty-parm-tags
    
    p   >>
        Does nothing (only there for compatibility with older versions of «yml2c»).
        >>

    h3 > -n NORMALIZATION, --normalization=NORMALIZATION

    p   >>
        Unicode normalization (none, NFD, NFKD, NFC, NFKC, FCD, default is NFC)
        >>

    h3 > -o FILE, --output=FILE

    p > Don't output to stdout, instead place output in file «FILE».

    h3 > -p, --parse-only

    p   >>
        ¬http://fdik.org/pyPEG Parse¬ only, then output ¬http://fdik.org/pyPEG/#pyAST pyAST¬
        as text to stdout. The pyAST is the
        ¬http://en.wikipedia.org/wiki/Abstract_syntax_tree Abstract Syntax Tree¬
        of the parsed script as ¬http://docs.python.org/library/stdtypes.html Python data structure¬.
        >>

    h3 > -V, --version

    p   >>
        Show version and ¬http://en.wikipedia.org/wiki/Copyleft Copyleft¬ information of
        the YML implementation.
        >>

    h3 > YML_PATH environment variable

    p   >>
        If «YML_PATH» is set to a colon separated ¬http://en.wikipedia.org/wiki/Path_(computing) path¬
        list (semicolon separated on Windows), then the
        `a href="features#including" code "include";` statement searches these paths.
        If «YML_PATH» is not set, the local directory «.» is searched for files.
        The system location for «.yml2» and «.ysl2» files is always searched afterwards.
        >>

    h2 id=processor > The YML/YSLT processor

    p   >>
        If you're processing more complex tasks, say, executing an YSLT script over some
        YML files, the YML/YSLT processor will possible fit better to your needs.
        >>

    p   >>
        To use it, enter «yml2proc -y» followed by the filename of your YSLT script,
        followed by the filenames of your YML input files:
        >>

    Code | % yml2proc -y myscript.ysl2 inputfile.yml2 > myscript.result

    p > The YML/YSLT processor requires the ¬http://codespeak.net/lxml/ lxml Python library¬.

    p > The YML/YSLT processor has the following options:

    h3 > -h, --help

    p > Show a help message and exit.

    h3 > -C, --old-syntax

    p > Use the syntax of YML 2 version 1.x (compatibility mode).

    h3 > -D, --emit-linenumbers

    p > Emit line numbers into the resulting XML for debugging purposes.

    h3 id=debug > --debug

    p   >>
        switch on debug tracing to stderr; this enables the functionality
        of the ¬yslt#debug debug(msg)¬ and ¬yslt#assert assert yml:assert(test, msg)¬ YSLT functions.
        >>

    h3 > -E ENCODING, --encoding ENCODING

    p > encoding of input files (default: UTF-8)

    h3 > -I INCLUDE_PATH, --include=INCLUDE_PATH

    p { "Precede " code > ¬features#ymlpath YML_PATH¬
      " by a colon separated " code "INCLUDE_PATH"; " (semicolon separated on Windows) to search for include files." }

    h3 > -m, --omit-empty-parm-tags
    
    p   >>
        Does nothing (only there for compatibility with older versions of «yml2proc»).
        >>

    h3 > -M, --empty-input-document

    p   >>
        Use an empty input document for YSLT or XSLT processing
        >>

    h3 > -n NORMALIZATION, --normalization=NORMALIZATION

    p   >>
        Unicode normalization (none, NFD, NFKD, NFC, NFKC, FCD, default is NFC)
        >>

    h3 > -o FILE, --output=FILE

    p > Don't output to stdout, instead place output in file «FILE».

    h3 > -p, --parse-only

    p   >>
        ¬http://fdik.org/pyPEG Parse¬ only, then output ¬http://fdik.org/pyPEG/#pyAST pyAST¬
        as text to stdout. The pyAST is the
        ¬http://en.wikipedia.org/wiki/Abstract_syntax_tree Abstract Syntax Tree¬
        of the parsed script as ¬http://docs.python.org/library/stdtypes.html Python data structure¬.
        >>

    h3 > -e XPATH, --xpath=XPATH

    p > Execute XPath expression XPATH and print result

    p   >>
        If you omit -y YSLTSCRIPT, then the result is printed. If you have an additional
        -y YSLTSCRIPT parameter, then your YSLT script is processing the result of the
        XPath expression only instead of processing the complete content of your input documents
        >>

    h3 > -P, --pretty
    
    p > Pretty print output adding whitespace

    h3 > -x, --xml
    
    p > Input documents are XML already, don't try to convert them from YML

    h3 > -y YSLTSCRIPT, --yslt=YSLTSCRIPT

    p > Execute YSLT script YSLTSCRIPT

    h3 > -X XSLTSCRIPT, --xslt=XSLTSCRIPT

    p > Execute XSLT script XSLTSCRIPT

    h3 > -d PARAMS, --paramdict=PARAMS

    p   >>
        Call XSLT or YSLT script with dictionary PARAMS as XPath parameters. PARAMS is a Python
        expression which evaluates to a Python dictionary.
        >>

    p > Sample:

    p code > yml2proc -y myscript -d 'dict(a=3, b="element[@name=2]")'

    h3 > -s STRINGPARAMS, --stringparamdict=STRINGPARAMS

    p   >>
        Call XSLT or YSLT script with dictionary STRINGPARAMS as string parameters. STRINGPARAMS
        is a Python expression which evaluates to a Python dictionary.
        >>  

    p > Sample:

    p code > yml2proc -y myscript -s 'dict(forename="Hans", name="Meier")'

    h3 > -Y, --xml2yml

    p > Convert XML to normalized YML. Equivalent to -xy xml2yml.ysl2

    h3 > -V, --version

    p   >>
        Show version and ¬http://en.wikipedia.org/wiki/Copyleft Copyleft¬ information of
        the YML implementation and exit.
        >>

    h3 > YML_PATH environment variable

    p   >>
        If «YML_PATH» is set to a colon separated ¬http://en.wikipedia.org/wiki/Path_(computing) path¬
        list (semicolon separated on Windows), then the
        `a href="features#including" code "include";` statement searches these paths.
        If «YML_PATH» is not set, the local directory «.» is searched for files.
        The system location for «.yml2» and «.ysl2» files is always searched afterwards.
        >>
    
    h2 id=recommended > Recommended external tools

    h3 > XSLT Processor

    p   >>
        If you want to use ¬yslt YSLT¬ with «yml2c» instead of «yml2proc», you additionally need an
        ¬http://www.w3.org/TR/xslt XSLT¬ processor.
        >>

    p > For that case, I recommend the ¬http://xmlsoft.org/XSLT/xsltproc2.html xsltproc tool¬.

    p > To process YSLT, just use a pipe with the «xsltproc» command:

    Code | % yml2c customer.ysl2 | xsltproc - customer.xml > customer.html

    h3 > XML Renicing Tool

    p   >>
        Sometimes, you may want to have more human readable output with «yml2c». So an
        ¬http://www.w3.org/XML/ XML¬ renicing tool can be helpful.
        >>

    p > The ¬http://xmlstar.sourceforge.net/ XMLStarlet¬ command line tool may be of interest.

    p > Have a look on its «fo» command.

    h3 > Windows tool chain

    p   >>
        See also the tips for a
        ¬http://www.auchdieserschwachsinnmussinsinternet.de/2009-03/#s1236859037 Windows tool chain¬
        for YML.
        >>

    h1 id=downloads > Downloads

    p > You can here ¬https://gitea.pep.foundation/fdik/yml2 download the newest release of YML¬.

    div id=bottom {
        a href="yslt" "<< back to YSLT" " "
        a href="#top" "^Top^" " "
        a href="toolchain.en.yhtml2" "(source)"
    }
}