Changing function signature of parse to accept the file content instead of a file
#1706
Conversation
…tead of a file This PR changes the way `parse` works (in a backwards compatible way). Instead of parsing a `File`, we parse the file contents (and a path). The reasoning behind this is that almost all language frontends currently need to read the file contents and we can harmonize this. This will also allow us to provide more common statistics about the parsing context in the future.
|
...ge-python/src/main/kotlin/de/fraunhofer/aisec/cpg/frontends/python/PythonLanguageFrontend.kt
Show resolved
Hide resolved
| if (path != null) { | ||
| path.absolute().toString() | ||
| } else { | ||
| "<unknown>" |
There was a problem hiding this comment.
To we want this constant for cpg consistency? Python recommends <string> see https://docs.python.org/3/library/functions.html#compile
There was a problem hiding this comment.
We are using ast.Parse and the default parameter here was <unknown>. See https://docs.python.org/3/library/ast.html#ast.parse
There was a problem hiding this comment.
Funny, that the ast.parse doc says:
Equivalent to compile(source, filename, mode, PyCF_ONLY_AST).
But I'm ok with the "unknown". This should probably a sane constant for all frontends if there is no file available?
There was a problem hiding this comment.
Funny, that the ast.parse doc says:
Equivalent to compile(source, filename, mode, PyCF_ONLY_AST).
But I'm ok with the "unknown". This should probably a sane constant for all frontends if there is no file available?
I was intentionally using the python specific value here, but I am fine with either way. Not sure if the SARIF standard defines something in this case, since we are using the location information from SARIF.
This PR changes the way
parseworks (in a backwards compatible way). Instead of parsing aFile, we parse the file contents (and a path). The reasoning behind this is that almost all language frontends currently need to read the file contents and we can harmonize this. This will also allow us to provide more common statistics about the parsing context in the future.Language frontends that can potentially leverage this: