TranslationFrameworkPage
From IndLinux
Contents |
Translation Check Framework
Overview
Translated User Interface content is often interspersed with errors that compromise the semantics, and overall quality of the results. Identification of these procedural errors and preventive checkpoints can mitigate their recurrance during the translation process and eventually be adopted as best practices. The checkpoints are focussed to be integrated during the translation process, such that erroneous results can be immediatly identified by the translator/reviewer and corrected.
Challenges
(information goes in here)
Probable Error Scenarios
- Translation Errors
- Language syntax (spelling etc.)
- Context
- Consistency
- Technical values (variables, commands, options, applications, system values)
- Keyboard shortcuts
- Reference mismatch
- Procedural Errors
- Language Addition/LINGUAS
- ChangeLogs/Commit Logs
- Submission and Branching
- File Checks (msgfmt, encoding)
- Caveats
- Brand Names
- Application Names
- Over-translation
Currently Available Solutions
- Manual Checks
- Limited Tools
Possible References
Framework Specifications
Broad Objectives
- Identifying Checkpoints
- Creating tools for these checkpoints
- Compliance by translators and reviewers
Identifying Checkpoints
| Test Summary | Input | Task | Output | Output Format | Special Notes |
|---|---|---|---|---|---|
| Handling of Reference Context (divided with 'pipe' in msgid) | The following input: 1. single .po file or, 2. multiple .po files, or 3. folder containing .po files | The script do the following steps: step 1: accept .po file(s)/folder with .po files, step 2: identify any pipe divided strings in the msgid, step 3: check if the corresponding msgstr contains the part before the pipe, step 4: if the part is contained, then error message should be generated and output of file details to be shown | error notification, filename, string number | Pipe Divided String error is the following: abc.po:2,4,5,6 | (notes go here) |
| Defined strings not to be translated | The following input: 1a. single .po file, or 1b. multiple .po file, or 1c. folder containing .po files, and 2. data file containing the defined strings | The script should do the following steps: step 1: accept .po file(s)/folder with .po files, step 2: script should automatically load the datafile containing the strings not to be translated, step 3: check if the msgid contains any of the strings from the data file step 4: if the msgstr does not contain the defined string, then error message should be generated and output of file details to be shown | error notification, filename, string number | String definition error is the following: abc.po:2,4,5,6 | Notes: 1. The datafile should be customizable and editable 2. User can define their own datafile and use it along with the default data file 3. User can choose to use their own datafile only and ignore the default datafile, 4. The datafile would require regular updates |
| Keyboard Shortcuts | The following input: 1. single .po file or, 2. multiple .po files, or 3. folder containing .po files | The script should do the following steps: step 1: accept .po file(s)/folder with .po files, step 2: script should first check if any of the strings do not contain the shortcuts in the msgstr, step 3: strings without the shortcut entries would need to be marked as (error message 1) step 4: the script should have an option for the user to specify if he/she wants to check english/non-english shortcuts step 5: if the msgstr contains shortcuts, then the script has to check if the shortcuts are in english or not, as per the option specified by the user step 6: if the shortcuts are not in the correct specification then an additional error message (error message 2) should be shown | error notification, filename, string number | (error message 1) Keyboard shortcut missing in the following: abc.po:2,4,5,6 (error message 2) Keyboard shortcuts are not in the prescribed language: abc.po:3,7,9,11 | (notes go here) |
Feedback
- (Add here)
