100% match: segment in document and TM are identical, no further conditions

Context match: segment in document and TM are identical, the document structure level matches, the previous segment in source is the same and has the same translation as in previous document

Perfect Match: segment in document and TM are identical, surrounding segments in source and target are identical (before and behind afaik), document structure matches

Firstly, thanks for your efforts and time.

Your comment is brief and kind but not clear.

100% match is too easy, so does not needed any more wordings.
Others, two matches, you have introduced "structure" and "previous" or "surrounding" concepts but, there is no such thing in my hand.
In short, I have a single separated sentence(or segment) in a document and a single separated sentence(or segment) in a TM, that is all. 

Document structure:

• Heading
• List item
• Paragraph
• Table cell

More document structure

• Line break
• Soft line break
• Section break
• Pagination
• Header/footer

Surrounding should be clear. These are the elements after segmentation (so segments) preceding and following your current segment.

Make a simple check:

This is my first sentence.
This is my second sentence.
This is my third sentence.
This is my fourth sentence.
This is my third sentence.
This is my fourth sentence.

Create 2 or 3 documents with this text. Format these segments in different way, making some of these heading in one document, but not in the second and try to translate.

Looking good
I'll try it out

Thanks again.

I have tested it.
But, nothing reasonable happened.

I have made a simple project using a docx and a fresh TM(no TUs within it).
And, I have finished translation, so, the TM is filled with the document's segments.

Now

I have made a simple project again using the same docx and the same TM (of course, it filled with TUs from the "previous" docx structure).
So, How's your idea ? It should displayed "PerfectMatch" all of them, right ?

Why is that ?
I guessed "100% Match" and You guessed "PerfectMatch" but, it shows "ContextMatch"es only.

attaching docx file showed at the first screenshot
TestingMSWord.zip
 

Let me test it for you and post the results here later on, when I have time.

sure
I understand that You do not have any kind of duty here.
I just say thanks again.

Enjoy

Kelly Edward

Your results are entirely as expected.  Perfect Match doesn't work that way and it's a feature only available to users with a Professional licence.  The way it works is that you would compare your file with a previously translated document in addition to your TM (although a TM is not necessary).  The main usecase for it has always been the "mid-project update" and the "leverage from a single bilingual document". It aligns the currently translated document's segments with those of the leveraged file, then recycles (copies) non-modified blocks of bilingual material into the new file, finally locking the recycled blocks and so preventing them from modification (optionally).  PerfectMatch does not disambiguate exact matches but reduces the scope of the translation (and subsequent review) to the difference between the current document and its previous version.

Kelly Edward

Your results are entirely as expected.  Perfect Match doesn't work that way and it's a feature only available to users with a Professional licence.  The way it works is that you would compare your file with a previously translated document in addition to your TM (although a TM is not necessary).  The main usecase for it has always been the "mid-project update" and the "leverage from a single bilingual document". It aligns the currently translated document's segments with those of the leveraged file, then recycles (copies) non-modified blocks of bilingual material into the new file, finally locking the recycled blocks and so preventing them from modification (optionally).  PerfectMatch does not disambiguate exact matches but reduces the scope of the translation (and subsequent review) to the difference between the current document and its previous version.