sciwrite-lint 0.2.0__py3-none-any.whl

sciwrite_lint/models.py

@@ -0,0 +1,139 @@
+"""Data models for sciwrite-lint."""
+
+from __future__ import annotations
+
+from typing import Any, Literal
+
+from pydantic import BaseModel, Field
+
+
+class Finding(BaseModel):
+    """A single issue found by a check."""
+
+    level: Literal["error", "warning", "info"]
+    rule_id: str  # check ID, e.g., "dangling-cite", "reference-exists"
+    message: str
+    file: str = ""
+    line: int | None = None
+    context: str = ""
+
+
+class CheckMeta(BaseModel):
+    """Metadata for a registered check."""
+
+    id: str  # "dangling-cite", "cross-section-consistency", etc.
+    severity: Literal["error", "warning", "info"]
+    category: Literal["manuscript", "reference-db", "local-llm"]
+    description: str
+
+
+# Legacy alias for backwards compatibility
+RuleMeta = CheckMeta
+
+
+class CheckResult(BaseModel):
+    """Result from running checks on a paper."""
+
+    checker: str  # e.g., "check", "verify"
+    paper: str  # e.g., "my_paper" or filename
+    findings: list[Finding] = Field(default_factory=list)
+
+    @property
+    def errors(self) -> list[Finding]:
+        return [f for f in self.findings if f.level == "error"]
+
+    @property
+    def warnings(self) -> list[Finding]:
+        return [f for f in self.findings if f.level == "warning"]
+
+    @property
+    def ok(self) -> bool:
+        return not self.errors
+
+
+class Citation(BaseModel):
+    """A citation extracted from a .tex file."""
+
+    key: str
+    raw_text: str
+    authors: list[str] = Field(default_factory=list)
+    title: str = ""
+    year: str = ""
+    venue: str = ""
+    doi: str = ""
+    url: str = ""
+    arxiv_id: str = ""  # e.g. "2310.01798"
+    pmid: str = ""  # e.g. "12345678"
+    pmc_id: str = ""  # e.g. "PMC1234567"
+    isbn: str = ""  # e.g. "9780226458113"
+    lccn: str = ""  # e.g. "2019012345"
+    source_paper: str = ""
+    bib_format: str = ""  # "simple", "natbib", or "bib"
+    entry_type: str = ""  # BibTeX entry type: "article", "misc", etc.
+
+    # Local source
+    local_status: Literal["pdf", "md", "none"] = "none"
+    local_path: str = ""
+
+    # API verification
+    api_match: Literal[
+        "",
+        "verified",
+        "mismatch",
+        "not_found",
+        "skipped",
+        "web_verified",
+        "web_dead",
+        "manual",
+    ] = ""
+    api_source: str = ""
+    api_data: dict = Field(default_factory=dict)
+    issues: list[str] = Field(default_factory=list)
+
+    # Tier (computed from metadata)
+    tier: Literal["T1", "T2", "T3", ""] = ""
+
+
+class CitationMetadata(BaseModel):
+    """Persistent per-citation verification record stored as JSON."""
+
+    key: str
+    verified_date: str = ""
+    api_source: str = ""  # "crossref", "openalex", "semantic_scholar"
+    api_match: Literal[
+        "",
+        "verified",
+        "mismatch",
+        "not_found",
+        "skipped",
+        "web_verified",
+        "web_dead",
+        "manual",
+    ] = ""
+
+    # Canonical data from API (what the world says)
+    canonical: dict[str, Any] = Field(default_factory=dict)
+
+    # What our .tex bibitem says
+    bibitem: dict[str, Any] = Field(default_factory=dict)
+
+    # Access status
+    access: dict[str, Any] = Field(default_factory=dict)
+
+    mismatches: list[str] = Field(default_factory=list)
+    issues: list[str] = Field(default_factory=list)
+
+    # Manual override
+    manual_override: dict[str, Any] = Field(default_factory=dict)
+
+
+class SectionInfo(BaseModel):
+    """A section in a LaTeX document."""
+
+    label: str  # e.g., "sec:intro" (empty if no label)
+    title: str
+    depth: int  # 0=section, 1=subsection, 2=subsubsection
+    start_line: int
+    end_line: int
+    word_count: int
+    cite_count: int = 0

sciwrite_lint/pdf/__init__.py

@@ -0,0 +1 @@
+"""PDF acquisition and parsing: GROBID and download utilities."""