65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179 | def generate(data_dir: str) -> list[dict] | None:
"""Collect paper citation counts and write results + history."""
data_path = Path(data_dir)
artifacts_path = data_path / "assets" / "data" / "artifacts.json"
build_dir = data_path / "_build"
build_dir.mkdir(parents=True, exist_ok=True)
out_path = build_dir / "paper_citations.json"
if not artifacts_path.exists():
logger.error("artifacts.json not found at %s", artifacts_path)
return None
artifacts: list[dict] = load_json(artifacts_path)
logger.info("Loaded %d artifacts", len(artifacts))
# Deduplicate by normalized title
seen_titles: set[str] = set()
unique: list[dict] = []
for a in artifacts:
norm = normalize_title(a.get("title", ""))
if norm and norm not in seen_titles:
seen_titles.add(norm)
unique.append(a)
logger.info("%d unique papers to process", len(unique))
session = create_session()
s2_disabled = os.environ.get("DISABLE_SEMANTIC_SCHOLAR", "").strip() == "1"
s2_timeout_failures = 0
entries: list[dict] = []
cached_count = 0
fetched_count = 0
for i, artifact in enumerate(unique):
title = artifact.get("title", "")
norm = normalize_title(title)
doi = _extract_paper_doi(artifact.get("paper_url"))
# Try cache first (keyed by DOI or normalized title)
cache_k = _cache_key(doi) if doi else _cache_key(norm)
cached = read_cache(CACHE_DIR, cache_k, CACHE_TTL, CACHE_NS)
if cached is not _MISSING:
entries.append(cached)
cached_count += 1
continue
# Lookup via APIs
openalex_count: int | None = None
s2_count: int | None = None
source = ""
openalex_id = ""
if doi:
time.sleep(_OPENALEX_DELAY)
oa = _openalex_lookup(doi, session)
if oa:
openalex_count = oa["cited_by_count"]
openalex_id = oa["openalex_id"]
source = "openalex_doi"
if not s2_disabled:
time.sleep(_S2_DELAY)
s2_count = _s2_lookup(doi, session)
if s2_count is None and s2_timeout_failures < _S2_MAX_TIMEOUT_FAILURES:
s2_timeout_failures += 1
if s2_timeout_failures >= _S2_MAX_TIMEOUT_FAILURES:
logger.warning("Disabling Semantic Scholar after %d timeout failures", s2_timeout_failures)
s2_disabled = True
else:
# No DOI — fall back to OpenAlex title search
time.sleep(_OPENALEX_DELAY)
oa = _openalex_title_search(title, session)
if oa:
openalex_count = oa["cited_by_count"]
openalex_id = oa["openalex_id"]
source = "openalex_title"
cited_by = best_citation_count(openalex_count, s2_count)
entry = {
"title": title,
"conference": artifact.get("conference", ""),
"year": artifact.get("year", 0),
"category": artifact.get("category", ""),
"ae_paper": True,
"paper_doi": doi,
"openalex_id": openalex_id,
"cited_by_count": cited_by,
"citations_openalex": openalex_count,
"citations_semantic_scholar": s2_count,
"source": source,
}
entries.append(entry)
write_cache(CACHE_DIR, cache_k, entry, CACHE_NS)
fetched_count += 1
if (i + 1) % 100 == 0:
logger.info("Progress: %d/%d (cached=%d, fetched=%d)", i + 1, len(unique), cached_count, fetched_count)
logger.info(
"Done: %d entries (cached=%d, fetched=%d, with_citations=%d)",
len(entries),
cached_count,
fetched_count,
sum(1 for e in entries if isinstance(e.get("cited_by_count"), int) and e["cited_by_count"] > 0),
)
save_json(out_path, entries)
logger.info("Wrote %s", out_path)
# ── History tracking ─────────────────────────────────────────────────
_update_history(entries, build_dir)
return entries
|