popp

Features

The preprocessor allows your translated text to reference other items of translated text, in a way that should work within any .PO editor. The use of $include statements is also supported, and conditional directives might be added in future.

See the GNU gettext documentation for information about PO files.

For example, you can have .PO source files like this:

msgid "ProductName short"
msgstr "POpp"

msgid "Welcome text"
msgstr "Congratulations on your download of {id:productname short}"

and it can be converted automatically to this:

msgid "ProductName short"
msgstr "POpp"

msgid "Welcome text"
msgstr "Congratulations on your download of POpp"

Current version:

• The Windows executable can be downloaded here (v0.2.0)
  • md5: 51f4071976faf82f7cbcd8cf9325fe02 *popp.exe
• I'm still in the process of trying popp in a real project environment, so for now assume it's only been tested with the automatic test cases, and take care.

Documentation:

.PO msgstrs can reference other msgstr values using a curly brace notation, for example {id:ProductName_short} will be expanded to the msgstr which has the msgid of ProductName_short.

Usage:

popp [options] source.popp [dest.po]

Brace notation:

• {id:msgid} or {id:msgid-msgctxt}
• references can be escaped with a backslash, e.g. \{id:msgid} is ignored.
  

WARNING: Plural forms are not supported, the file can still be processed, however lines begining with "msgstr[n]" will not have their content expanded, and plural forms cannot be referenced with the brace notation.

$include notation

The following include statements are all valid notation

$include "file.po"

# $include "file.po"
#.$include "file.po"

The .po hash-comment notations can be used if the file must be editable or parsable by other .po tools before being processed by popp.

Output, and pipes

Output files are written in UTF-8. If your source languages use unicode characters that your shell can't display, then avoid using pipes and stick with specifying an input file and an output file. Or proceed very carefully, as you risk having the unicode characters mapped to what the shell can handle.

Options:

• -nl, --nLF

  Use LF for newlines

• -nc, --nCRLF

  Use CRLF for newlines

• -ns, --nSource

  [Default] Determines LF or CRLF for newlines by what the source file uses

• -s, --sensitive, --casesensitive

  The msgids inside curly-brace-references are matched case-insensitively by default, the --sensitive option expands case-sensitive matches only.

• -c, --count

  Returns the number of references contained in the source file, regardless of whether the references are valid and can be expanded. No output file is written. Can be used as a second pass to confirm all references have been expanded (none were misspelled etc), but

  WARNING: Plural forms are not supported and references contained in plural-form msgstrs are not counted.

• -i [path], --includeDirectory [path]

  Adds a directory to the end of the search path used to locate files specified by $include directives.

• -q, --quiet

  Suppresses console error messages and info messages

Returns:

• 0 - Success
• 1 - fatal error - invalid arguments or file permissions.
• 2 - fatal error - non-specific.
• 3 - one or more warnings or non-fatal errors occurred.
• less than 0 - Success, but not all references could be expanded, the negative return value indicates how may references were found that could not be expanded.