Draft: moving over citation_exists.py requirement #7
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…est; also added eyecite useage for LLM output parsing
nrfulton
requested changes
Nov 4, 2025
Comment on lines
+36
to
+75
| # might not be needed | ||
| # def ensure_list_of_dicts(obj: Any) -> list[dict]: | ||
| # """ | ||
| # Normalize any JSON-like object into a list of dictionaries. | ||
|
|
||
| # Accepts: | ||
| # - A JSON string (object or array) | ||
| # - A single dict | ||
| # - A list of dicts | ||
|
|
||
| # Args: | ||
| # obj: Any data type, ideally something that can unpacked into a dictionary | ||
|
|
||
| # Returns: | ||
| # The unpacked object in list of dictionary form or raises an error. | ||
| # """ | ||
| # # JSON string | ||
| # if isinstance(obj, str): | ||
| # try: | ||
| # obj = json.loads(obj) | ||
| # except json.JSONDecodeError as e: | ||
| # raise ValueError(f"Invalid JSON string: {e!s}") | ||
|
|
||
| # # Single dict | ||
| # if isinstance(obj, dict): | ||
| # return [obj] | ||
|
|
||
| # # List of dicts | ||
| # if isinstance(obj, list): | ||
| # if all(isinstance(item, dict) for item in obj): | ||
| # return obj | ||
| # else: | ||
| # raise ValueError("List contains non-dictionary elements") | ||
|
|
||
| # raise TypeError(f"Unsupported metadata format: {type(obj)}") | ||
|
|
||
| # alternatively: | ||
| # should this take in last_output instead of the whole context? | ||
| # get case name: take LLM output and extract case name --> a string which you get from ctx.last_output() is the input | ||
| # so the argument should be ctx.last_output.value: str |
There was a problem hiding this comment.
- What was the purpose of this?
- Remove commented out code.
Comment on lines
+89
to
+96
| # install hyperscan if not already installed | ||
| # !pip install hyperscan | ||
| # tokenizer = HyperscanTokenizer(cache_dir=".test_cache") | ||
| # citations = get_citations(cleaned_text, tokenizer=tokenizer) | ||
|
|
||
| # or this? | ||
| # cleaned_text = clean_text(text, ["html", "all_whitespace"]) | ||
| # citations = get_citations(cleaned_text) |
Comment on lines
+103
to
+108
| plaintiff = citation.metadata.get("plaintiff") | ||
| defendant = citation.metadata.get("defendant") | ||
| if plaintiff and defendant: | ||
| case_names.add(f"{plaintiff} v. {defendant}") | ||
| # name = citation.metadata['plaintiff'] + " v. " + citation.metadata['defendant'] | ||
| # case_names.add(name) |
There was a problem hiding this comment.
- What's the purpose of this?
- Are these actually canonical names/references?
- What happens if you don't have a plaintiff and defendent? Can that ever happen? If not -> assert. If yes -> handle exceptiona lcases.
Comment on lines
+163
to
+173
| # check if this code chunk is right later | ||
| # db_names = {normalize_case_name(c["name"]) for c in case_metadata if "name" in c} | ||
| # db_abbrevs = { | ||
| # normalize_case_name(c["name_abbreviation"]) for c in case_metadata if "name_abbreviation" in c | ||
| # } | ||
|
|
||
| # for name in normalized_output_names: | ||
| # if name not in db_names and name not in db_abbrevs: | ||
| # return ValidationResult(False, reason=f"Case '{name}' not found in database") | ||
|
|
||
| # return ValidationResult(True, reason="All case names found in database") |
There was a problem hiding this comment.
remove commented out code.
| case_names = extract_case_names(ctx) | ||
|
|
||
| if not case_names or not isinstance(case_names, list[str]): | ||
| return ValidationResult(False, reason="No case names provided in output") |
There was a problem hiding this comment.
I think this should probably return True. The reason is good.
Comment on lines
+148
to
+152
| for case in case_metadata: | ||
| if 'name' in case: | ||
| case_names.add(normalize_case_name(case['name'])) | ||
| if 'name_abbreviation' in case: | ||
| case_name_abb.add(normalize_case_name(case['name_abbreviation'])) |
There was a problem hiding this comment.
This approach seems like it will make a lot of errors.
What about cases where the case name isn't verbatim? E.g., sometimes if there is a large set of parties on one side or the other there will be an abbreviation of that in the cite. State names are often also different in the formal cite vs how they're cited inline.
Additionally, there is a lot of string manipulation here that I think can be streamlined or done in a more principles way. Should we implement this as something like likely_equivalent(c1, c2) where c1 and c2 are eyecite citation objects.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Draft: moving over citation_exists.py requirement and corresponding test; also added eyecite useage for LLM output parsing