stor-toolchain/toolchain/writers/tmx_creator.py

import argparse
import logging
import time
from jinja2 import Template

logger = logging.getLogger(__name__)

class TmxCreator:

    DATE_FORMAT = "%Y%m%dT%H%M%SZ" # To meet ISO 8601 and as recommended by 1.4b
    REQUIRED_ARGS = [
        "adminlang",
        "datatype",
        "o_tmf",
        "segtype",
        "srclang",
        "tgtlang",
    ]

    def create(self, input_path_template, input_path_src, input_path_tgt, output_path, additional_args={}):
        logger.info("Creating TMX file from {0} and {1} to {2}.".format(input_path_src, input_path_tgt, output_path))
        logger.debug("Creating TMX file to {0} using template {1} and additional args {2}.".format(output_path, input_path_template, additional_args))
        with open(input_path_template) as template_file, open(input_path_src) as contents_src,\
                open(input_path_tgt) as contents_tgt, open(output_path, "w") as output_file:
            self.create_tmx(template_file.read(), contents_src, contents_tgt, output_file, additional_args)


    def create_tmx(self, template_input, input_src, input_tgt, output, additional_args):
        template = Template(template_input, autoescape=True)
        creation_date = self.make_creation_date()
        input_src = map(str.strip, input_src)
        input_tgt = map(str.strip, input_tgt)
        tus=zip(input_src, input_tgt)

        args = {
            "creation_date" : creation_date,
            "tus" : tus,
        }
        for k, v in additional_args.items():
            args[k] = v

        for required_arg in self.REQUIRED_ARGS:
            if not required_arg in args:
                logger.warning("Attribute {0} required by TMX 1.4b not given.".format(required_arg))

        output.write(template.render(args))


    def make_creation_date(self):
        return time.strftime(self.DATE_FORMAT)


if __name__ == "__main__":
    argparser = argparse.ArgumentParser()
    argparser.add_argument("template_path", help="path to template file")
    argparser.add_argument("input_path_src", help="path to input file of source language")
    argparser.add_argument("input_path_tgt", help="path to input file of target language")
    argparser.add_argument("output_path", help="path to output TMX file")
    argparser.add_argument("adminlang", help="language code of administrative language")
    argparser.add_argument("datatype", help="type of data contained")
    argparser.add_argument("disclaimer", help="disclaimer text")
    argparser.add_argument("distributor", help="resource distributor")
    argparser.add_argument("licence", help="licence name")
    argparser.add_argument("o_tmf", help="original translation memory format")
    argparser.add_argument("segtype", help="segmentation type")
    argparser.add_argument("srclang", help="language code of source language")
    argparser.add_argument("tgtlang", help="language code of target language")
    args = argparser.parse_args()

    additional_args = {
        "adminlang" : args.adminlang,
        "datatype" : args.datatype,
        "disclaimer" : args.disclaimer,
        "distributor" : args.distributor,
        "licence" : args.licence,
        "o_tmf" : args.o_tmf,
        "segtype" : args.segtype,
        "srclang" : args.srclang,
        "tgtlang" : args.tgtlang,
    }

    TmxCreator().create(args.template_path, args.input_path_src, args.input_path_tgt, args.output_path, additional_args)