This article shows the main analysis points related to the QA Check functionality.
1. QA Check Tool Structure
The QA Check tool serves as a resource for linguists and project managers to verify potential inconsistencies in their translations. It enables quick identification of key areas for improvement and offers suggested fixes provided by the BWX Editor.
- To utilize the QA feature, access the BWX Editor by clicking on the Work Unit(s) for which you intend to run a QA check.
⚠️ The QA Check functionality can also be employed for Joined Work Units, allowing the selection between "Run for Current" file or "Run for All" files.
- Click on the Show/Hide Menu option and navigate to the QA Issues tab
- Following the QA Check on this tab, the warnings will be displayed in a list format, as the image below:
Select the Filters option to set up the specific points you'd like the QA to target.
- Once you've chosen the desired QA settings to execute, simply click on Apply and the QA Check will initiate.
- To conceal the ignored issues, activate the Hide ignored issues option from the QA options list. Then, click on the Eye Icon (Ignore issue) corresponding to the warning you wish to disregard. Subsequently, the warning will no longer be visible on the Warning list.
⚠️ Please note that ignored issues will not be automatically ignored in subsequent workflows. If a specific issue needs to be disregarded in all of the file's workflows, the professional handling that particular workflow must also ignore the issue. The configuration is tied to the work unit(s), meaning that if you access the same workflow with a different user, the ignored issues will remain ignored and the visible issues will stay visible.
To be redirected to the analysis segment, click on
Go to segment
2. Analysis Points
Here are concise explanations highlighting the primary quality checkpoints recommended by the QA process. These points serve as key indicators to ensure the quality and accuracy of the work being evaluated.
2.1.1 Warn if narrow non-breaking spaces mismatch
Checks if there are any differences of narrow non-breaking spaces between source and target.
2.1.2 Warn if non-breaking spaces mismatch
Checks if there are any differences of non-breaking spaces between source and target.
2.1.3 Warn if glossary mismatch
Checks if there are terms in the Glossary that were translated incorrectly.
2.1.4 Warn on doubled words
Checks if there are duplicated words in the translation.
In the image above, the QA pointed out with a SUSPECT PATTERN flag that the word "State" was repeated in the target.
2.1.5 Warn if a target segment is not empty when its source is empty
This option points to the segments whose target is not empty when the source is empty.
2.1.6 Warn if a target segment is empty when its source is not empty
Checks if the translated segments are empty when the corresponding source is not empty.
In the image above, the check issues an EMPTY TARGET flag indicating that the translated segment is empty, but the corresponding source is not empty.
2.1.7 Warn if repetitions are different
Checks if there are any repetitions in the source whose translations are different.
In the image above, the QA Check has issued a "Repetition Mismatch" flag signalizing that there are repetitions on the source content whose translations are different.
2.1.8 Warn if there are potential spelling errors
Checks if there are any potential spelling errors.
2.1.9 Warn if a target segment is the same as its source
Points out the segments in which the translation is identical to the source language.
2.1.10 Warn if whitespace mismatch
Signals the segments where the whitespaces are different between the source language and the target language.
2.2.1 Warn if there are tag differences between source and target segments
This option checks if the translation has the same tags as the source. The images below show some cases pointed out by the tool.
In the image above, the tool has issued a MISSING TAG flag to show that the linguist has not added a tag that should have been added to the target.
In this case, QA has issued an EXTRA TAG flag indicating that there is a tag in the target that is not in the source language.
2.2.2 Try to guess opening/closing types for placeholder tags
Enable this option to allow the system to attempt the identification of whether placeholder tags function as genuine opening or closing tags within XML files. When an opening or closing type is detected, it becomes part of the verification process for the opening/closing sequence.
2.2.3 Enforce strict order of the tags
This option checks if the tags are in the correct order.
In the example above, the QA has issued a SUSPECT TAG flag pointing out that there is an atypical tag sequence in the target.
2.3.1 Warn if a character is not included in the ISO-8859-1 character set encoding
This option checks the text characters against the ISO-8859-1 character set encoding.
The tool has issued an ALLOWED CHARACTERS flag pointing out that there is an inconsistency regarding the encoding of the character set. These checkpoints to a special character (Ξ) entered in the target.
2.3.2 Warn if some possibly corrupted characters are found in the target entry
This option points out special patterns that usually indicate a file with corrupted characters. For example, a UTF-8 file opens as ISO-8859-1, etc.
This check issued a SUSPECT PATTERN flag to point to the presence of a special symbol (Ãµ) entered in the target, indicating the possibility of corrupted characters.
⚠️This feature does not find all possible cases of corrupted characters, only some of the most frequent ones.
2.4.1 Warn if a target is longer than n chars
By default, this option addresses segments longer than 255 characters, but you have the flexibility to insert a different number if necessary. The character count is the basis for this limit and excludes tag elements.
In the image above, the tool emitted the TARGET LENGTH tag that indicates that the target has more than 255 characters.
2.4.2 Warn if a target is disproportionately longer than its source
This option points out any target segment that is disproportionately longer than its source text.
In the image above, the tool emitted the TARGET LENGTH tag to point out that the target is much larger than the source.
2.4.3 Warn if a target is disproportionately shorter than its source
This option points out any target segment that is disproportionately shorter than its source text.
2.4.4 Warn if a source or target text unit does not fit ITS storage size property
Set this option to check the byte length of the text units that have an ITS storage size property.
Read more about storage size here.
This option warns if some characters are breaking the formatting of the CSV. Set this option to check some issues such as additional semicolons (;), double-quotes (") and line breaks mismatches.
2.5.1 Warn if a target has extra double-quotes
When enabled, this option will warn when finding segments with extra double-quotes in comparison to the source content.
2.5.2 Warn if a target has extra semicolons
When enabled, this option will warn when finding segments with extra semicolons in comparison to the source content.
2.5.3 Warn if a target has line breaks mismatches
When enabled, this option will warn when finding segments with missing or additional line breaks in comparison to the source content.
⚠️ In case of any false warnings, the linguist can ignore these QA Checks. They are not mandatory to deliver or complete any workflow. The QA Tool will not prevent the user from completing a task if the user decides to ignore any or all warnings. The QA Tool will only provide the user with some guidance on possible issues within the translated content so the user can make the required amendments.
If any questions were not answered in this article, please don't hesitate to contact our support team at firstname.lastname@example.org.