Again, this will be needed for eventually distinguishing `glyph`
from `[glyph]` or `@CLASS`, which makes a semantic difference
when building GPOS type 2 tables.
This will be needed for implementing class-based kerning. According
to the OpenType Feature File specification, a different table needs to
be built for `glyph` versus `[glyph]`.
Before this fix, the parser failed to process statements of the form
`lookupflag MarkAttachmentType @GLYPHCLASS;` when no other flag
were set after the glyph class. However, statements like
`lookupflag MarkAttachmentType @GLYPHCLASS RightToLeft;` were
getting recognized perfectly fine.
No output is generated yet, this change is just on the parser.
The OpenType Feature File specification is surprisingly vague about
the exact syntax of chaining contextual positioning rules, so I expect
that we will have to iterate on this parser. However, the test case
in parser_test.py gets recognized by `makeotf`, so the current
implementation is unlikely to be completely wrong.
While not really documented in the OpenType Feature File specification,
the AFDKO makeotf tool handles the `markClass` statement this way.
Also, some examples in the specification seem to imply this semantics.
The class names of tree nodes for substitution and positioning rules
are now consistent with `builder.py`, which in turn is consistent with
`otTables.py`.
However, not sure how to build the otTables object graph for emitting
GPOS tables with device values; the current code thus silently strips
off any device values. Left a TODO comment for implementing this.
Also add ';' to some langaugesystem test cases. This makes the
snippets syntactically valid. The parser is still expected to
reject them for other reasons, just as before this change.
This simplifies the public API to the library. For clients, it does
not matter which exact component was detecting an error. And we will
soon have more components; there would be little point in declaring
CompilerError, TableBuilderError, and so forth.
In the new version, the parse_languagesystem_() function returns
an ast.LanguageSystemStatement, which then gets appended to the
current list of statements by the caller. Before this change,
the parse_languagesystem_() function would always append the
newly parsed statement to the list of top-level statements.
In practice, it does not matter because `languagesystem` is
only valid at the top level, but there is no good reason why
the parser method would have to be different from all other
constructs.
We will soon support additional blocks beyond `feature`,
and keeping this refactoring separate from new functionality
makes it easier for code reviewers to follow the changes.
