[MU3 Backend] ENG-67, 69, 73, 74, 83: Staff Text inference clean-up #8678

iveshenry18 · 2021-07-21T07:49:06Z

Resolves:

ENG-67: "From SHOW" text is rendered as Staff Text
ENG-69: "This music has been legally downloaded..." occasionally renders as Staff Text
ENG-73: Some scores with inferred credits also have normal credits
ENG-74: Credit inference regex can be more flexible
ENG-83: Header VBox formatting needs improvement

This PR includes commits for a set of similar issues regarding incorrectly exported staff text (<direction> elements) that should be header text.

See individual commit messages for detailed descriptions of the fixes.

I signed CLA
I made sure the code in the PR follows the coding rules
I made sure the code compiles on my machine
I made sure there are no unnecessary changes in the code
I made sure the title of the PR reflects the core meaning of the issue you are solving
I made sure the commit message(s) contain a description and answer the question "Why do those changes fix that particular issue?" or "Why are those changes really necessary as improvements?"
I made sure the commit message title starts with "fix #424242:" if there is a related issue
I created the test (mtest, vtest, script test) to verify the changes I made

Similar to credits, sometimes Dolet exports source information ("from ABC") as direction words near the beginning of the piece. This commit adds a function for inferring these and adding them to the header as a subtitle.

Dolet also sometimes incorrectly exports footers as direction words, particularly "This music has been legally downloaded . Do not photocopy." This commit throws out staff text with these contents.

This commit refines the regex for inferring credit directions. It ultimately removes requirements for names (the part after the word "by") due to the inordinate difficulty of writing strict regex that catches all possible characters included in names. To mitigate false positives, it clarifies/restricts the regex for the part before the word "by". This compromise decreases the number of false negatives without significantly increasing false positives.

Some scores specify header text both in the <identification> element and in the incorrectly-exported staff text. The latter tends to have more information, so this commit hides the former when the latter is present to avoid redundancy.

Due to inferred (mis-exported) credits not having valid positioning information, there were occasionally collisions with the title text. This commit adds a reformatting function that runs at the end of the import (if any inferred header text is present) that prevents these vertical collisions.

This commit extends the cases covered by inferred subtitles. Also, it extends this functionality to apply to Titles—some titles are multi- line, where the second line is really a subtitle. This commit adds a function to infer these, extract them from the title, and add them as the subtitle.

This commit changes the vertical padding to 1.2 * the line height when reformatting a header box with inferred credits or subtitles.

This commit adds credit inferring to the title text (similar to what already exists for subtitles). It also adds and refines the inference cases for both credits and subtitles.

Similar to credits, sometimes Dolet exports source information ("from ABC") as direction words near the beginning of the piece. This commit adds a function for inferring these and adding them to the header as a subtitle. Duplicate of musescore#8678, part 1

Dolet also sometimes incorrectly exports footers as direction words, particularly "This music has been legally downloaded . Do not photocopy." This commit throws out staff text with these contents. Duplicate of musescore#8678, part 2

This commit refines the regex for inferring credit directions. It ultimately removes requirements for names (the part after the word "by") due to the inordinate difficulty of writing strict regex that catches all possible characters included in names. To mitigate false positives, it clarifies/restricts the regex for the part before the word "by". This compromise decreases the number of false negatives without significantly increasing false positives. Duplicate of musescore#8678, part 3

Some scores specify header text both in the <identification> element and in the incorrectly-exported staff text. The latter tends to have more information, so this commit hides the former when the latter is present to avoid redundancy. Duplicate of musescore#8678, part 4

Due to inferred (mis-exported) credits not having valid positioning information, there were occasionally collisions with the title text. This commit adds a reformatting function that runs at the end of the import (if any inferred header text is present) that prevents these vertical collisions. Duplicate of musescore#8678, part 5

This commit extends the cases covered by inferred subtitles. Also, it extends this functionality to apply to Titles—some titles are multi- line, where the second line is really a subtitle. This commit adds a function to infer these, extract them from the title, and add them as the subtitle. Duplicate of musescore#8678, part 6

This commit changes the vertical padding to 1.2 * the line height when reformatting a header box with inferred credits or subtitles. Duplicate of musescore#8678, part 7

This commit adds credit inferring to the title text (similar to what already exists for subtitles). It also adds and refines the inference cases for both credits and subtitles. Duplicate of musescore#8678, part 8