If a segment has been translated, you will no longer, by default, see a 100% match when you open the 'Translation' panel.

1. A translated segment is selected. Even though it is the same text as shown in the 'Translation Result' window, it is not a 100% match.
2. This is because the 'Translation Source' is set to 'Selected Text' (i.e. the default setting), which means that we are using the (translated) text in the document as the source segment.
3. However, our source language is 'English (United States)'.

Given that the text in the document is now German, it would make no sense if we got a 100% match in an English translation memory.

In fact, if we did see a 100% match in this case, we should be worried about the integrity of our translation memory.

If you wish to see the original English-to-German translation memory hits, simply change the 'Translation Source' to 'Source Text'. This will perform a match against the original source (English) segment instead of the translated (German) segment.

1. Segments that have special characters within them cannot get a 100% match. This is because the one2edit™ translation memory cannot store these special characters (it only contains plain text).
2. When you open the 'Translation Memory Info' window (click the 'i' button next to a translation result), you will see that this special character is not contained therein. This means that it will not be a 100% match.

Each InDesign special character has a hex value that can be detected by a regular expression. Therefore, you can fine-tune your segmentation rules to detect such special characters, breaking them off into their own segments.

1. The styling on this heading uses an 'uppercase' font face, even when lower-case characters are entered. However, the lower-case characters are still stored in the document.
2. An upper-case character is different to a lower-case character. When you open the 'Translation Memory Info' window (click the 'i' next to a translation result), you see that the document source text differs in capitalization from the translation-memory source text.
3. Therefore, only a 60% match is found due to these differences in text.

Keep in mind, that the translation memory is case sensitive, and that such styling settings can be a reason for not finding a 100% translation match.

NOTE:
Styling is not stored in the plain-text translation memory. Therefore, local styling (e.g. bold text) can also cause a pre-translated segment to not be marked as 'Done'. This is because, while the characters may be a 100% match, a human intervention is still required to apply the local styling to the correct text in the translation.