We ❤️ shell scripts! To package and assemble the binary artifact run the following command:
$ ./mvnw clean package
This will execute the assemble goal from the appassembler plugin, as well as generate the required .class files.
Two binaries will be generated. One for Unix systems and a .bat file for Windows systems.
On Linux or MacOS use this:
$ ./target/appassembler/bin/i18n out
On Windows use this:
$ .\target\appassembler\bin\i18n.bat out
On Linux or MacOS use this:
$ ./target/appassembler/bin/i18n in
On Windows use this:
$ .\target\appassembler\bin\i18n.bat in
Or you can just run the main method in your IDE.
** Be sure to run from the root directory with ./target/appassembler/bin/i18n and not navigating into the folder and running with ./i18n, it will get confused about filepaths!**
You need to specify a subcommand.
| Commands | Effect |
|---|---|
| ./i18n in | Parse incoming i18n files reviewed by helpdesk |
| ./i18n out | Generate excel sheets for i18n purposes to share with helpdesk |
New json files are put in src/main/resources/new. The old ones with which the program will compare the updated ones should go in the src/main/resources/old folder.
There should be a consensus on how and where these shall be preserved... Maybe after processing the incoming the 'new' ones can be put in the old folder and the old ones thrown out manually by the dev. The date metadata can come in handy here.
For now they are placed in target/generated-i18n-files/. There you will find labelsForReview.xlsx & labelsForReview.xls which are suitable to send out to helpdesk. They are identical in content.
There's a directory called target/generated-i18n-files/csv/ where you will find the intermediate csv's. You can use these to check which labels were changed, which were not yet translated and which were newly added. It's these sets that are combined in combinedLabelsForReviewNL/FR.csv.
There you will also find labelsNotSelectedForReviewFR/NL.csv which should be kept. These will be used for merging the reviewed labels that come back from helpdesk. The same dev will have to do the outgoing operation and process the incoming in order to have these complementary labels, or at least share them with the dev that will do the incoming operation.
I've pointed the parsing programming to src/main/resources/labelsForReview.xls or .xlsx, either one is fine.
Put the reviewed excel file AND both labelsNotSelectedForReviewFR/NL.csv files in src/main/resources/ and run the parser to generate incremented json for each language.
They will be exported to .json files in target/generated-i18n-files/incremented-json/.