rxiv-maker 1.16.7__tar.gz → 1.17.0__tar.gz

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/CHANGELOG.md

@@ -7,6 +7,123 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.17.0] - 2025-12-22
+
+### Added
+
+- **PDF Splitting Feature**: New `--split-si` flag for `rxiv pdf` command
+  - Automatically splits generated PDFs into main manuscript and supplementary information sections
+  - Auto-detects SI markers ("Supplementary Information", "Supplementary Material", "Supporting Information", etc.)
+  - Generates properly named split files: `{year}__{author}_et_al__rxiv__main.pdf` and `{year}__{author}_et_al__rxiv__si.pdf`
+  - Places split files in MANUSCRIPT directory with consistent naming convention
+  - Addresses common journal submission requirements for separate main and SI files
+- **Color-Coded DOCX References**: Visual distinction for different reference types
+  - Main figures: bright green highlighting
+  - Main tables: blue highlighting
+  - Supplementary elements (figures, tables, notes): turquoise highlighting
+  - Citations: yellow highlighting
+  - Equations: violet highlighting
+  - Improves readability and helps track different element types during review
+- **Citation Range Formatting**: Automatic formatting of adjacent citations
+  - Formats `[4][5][6]` as `[4-6]` for cleaner presentation
+  - Handles both adjacent bracket citations and comma-separated citations
+  - Maintains readability while reducing visual clutter
+- **DOCX Configuration Options**: New manuscript config options
+  - `docx.hide_si`: Control supplementary information visibility (default: false)
+  - `docx.figures_at_end`: Place main figures at end before bibliography (default: false)
+  - Clearer semantics than previous `add_si` option
+  - Supports different journal formatting requirements
+- **Subscript/Superscript Support**: Chemical formulas and mathematical notation in DOCX
+  - Supports `~subscript~` and `^superscript^` markdown syntax
+  - Enables proper formatting of chemical formulas (H~2~O) and exponents (x^2^)
+
+### Fixed
+
+- **Citation Extraction Bug**: Citations between triple-backtick code blocks now extract correctly
+  - Fixed regex processing order to handle triple backticks before single backticks
+  - Prevents citations from being incorrectly marked as "protected code content"
+  - Resolves issue where `[@key]` citations remained unconverted in DOCX output
+  - Added regression tests to prevent future occurrences
+- **Bibliography Formatting**: Enhanced cleanup for malformed BibTeX entries
+  - HTML entity decoding: `é` → `é`, `á` → `á`, `…` → `…`
+  - Malformed name repair: fixes split author names (e.g., "Pé and Rez, Fernando" → "Pérez, Fernando")
+  - Brace cleanup: removes stray braces after Unicode characters
+  - Handles escaped HTML entities from BibTeX (`\&\#233` → `é`)
+  - Whitespace normalization for cleaner bibliography entries
+- **Table Caption Formatting**: Supplementary tables now show proper labels
+  - Displays "Supp. Table SX." prefix with correct font sizing
+  - Preserves table label identifiers during processing
+  - Supports hyphens in table label names
+- **Comment Filtering**: Markdown metadata comments no longer appear in DOCX
+  - Filters comments starting with "note:", "comment:" (case-insensitive)
+  - Prevents manuscript-specific metadata from appearing in final output
+  - Improves multi-line comment parsing
+
+### Changed
+
+- **Figure Centering**: All figures now centered by default in DOCX exports
+  - Improves visual presentation and alignment with journal standards
+  - Consistent formatting across all figure types
+- **Citation Handling**: Citations treated as highlighted text instead of separate objects
+  - Maintains citation ranges in yellow highlighting
+  - Simplifies DOCX structure while preserving visual distinction
+- **PDF Spacing**: Optimized manuscript spacing for increased content density
+  - Tighter spacing between back matter section headers and text
+  - Improved overall visual hierarchy
+  - More content per page without sacrificing readability
+
+### Documentation
+
+- **Updated CLAUDE.md**: Enhanced development documentation
+  - Removed Docker/Podman engine references (moved to separate docker-rxiv-maker repository)
+  - Added PDF splitting feature documentation
+  - Clarified testing workflows and CI requirements
+
+## [1.16.8] - 2025-12-19
+
+### Added
+
+- **Automatic Poppler Installation**: Interactive installation prompt for DOCX export on macOS
+  - Detects missing poppler utilities when exporting to DOCX with PDF figures
+  - Offers automatic installation via Homebrew with user confirmation
+  - Shows clear installation instructions for Linux (apt) and other platforms
+  - Gracefully falls back to placeholders when poppler unavailable
+- **DOI Resolution Implementation**: `--resolve-dois` flag now fully functional for DOCX export
+  - Resolves missing DOIs using CrossRef API with intelligent title matching
+  - Cleans LaTeX commands from titles for accurate search results
+  - Validates matches using year information when available
+  - Handles network failures and timeouts gracefully
+  - Logs resolved DOIs for transparency
+- **Poppler Dependency Tracking**: Registered poppler in DependencyManager
+  - Added as system binary dependency with alternatives (pdftoppm, pdfinfo)
+  - Included in `rxiv check-installation` output
+  - Provides platform-specific installation hints
+
+### Fixed
+
+- **BST File Regex Bug**: Corrected bibliography style file generation to only modify `format.names`
+  - Previous regex matched both `format.names` and `format.full.names` functions
+  - Now specifically targets only `format.names` function using DOTALL flag
+  - Prevents corruption of citation labels in natbib
+  - Eliminates warning: "Found 2 format string patterns in .bst file"
+  - Raises error if unexpected matches found (defensive programming)
+- **PDF to Image Conversion**: Improved error handling for missing poppler utilities
+  - Distinguishes between poppler not installed, corrupted PDFs, and other errors
+  - Uses proper logging instead of print statements
+  - Re-raises poppler errors to allow CLI to offer installation
+  - Provides specific error messages for different failure types
+
+### Changed
+
+- **DOCX Export Experience**: Enhanced PDF figure embedding workflow
+  - Pre-flight check detects poppler availability before attempting conversion
+  - Caches poppler status to avoid repeated checks
+  - Shows helpful warnings with installation instructions
+  - PDF figures embed correctly when poppler is installed
+- **Homebrew Formula**: Added poppler as dependency
+  - Users installing via `brew install rxiv-maker` now get poppler automatically
+  - Ensures DOCX export works out of the box for Homebrew users
+
 ## [1.16.7] - 2025-12-18
 
 ### Changed

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rxiv-maker
-Version: 1.16.7
+Version: 1.17.0
 Summary: Write scientific preprints in Markdown. Generate publication-ready PDFs efficiently.
 Project-URL: Homepage, https://github.com/HenriquesLab/rxiv-maker
 Project-URL: Documentation, https://github.com/HenriquesLab/rxiv-maker#readme
@@ -56,6 +56,7 @@ Requires-Dist: mypy>=1.0; extra == 'dev'
 Requires-Dist: nbstripout>=0.7.1; extra == 'dev'
 Requires-Dist: nox>=2023.0.0; extra == 'dev'
 Requires-Dist: pre-commit>=4.2.0; extra == 'dev'
+Requires-Dist: pypdf2>=3.0.0; extra == 'dev'
 Requires-Dist: pytest-asyncio>=0.21.0; extra == 'dev'
 Requires-Dist: pytest-benchmark>=4.0.0; extra == 'dev'
 Requires-Dist: pytest-clarity>=1.0.0; extra == 'dev'

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/pyproject.toml

@@ -93,6 +93,8 @@ dev = [
     # Type checking
     "types-requests>=2.32.4.20250611",
     "types-PyYAML>=6.0.0",
+    # PDF utilities
+    "PyPDF2>=3.0.0",
     # Documentation and utilities
     "coverage[toml]>=7.0",
     "nbstripout>=0.7.1",

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/__version__.py

@@ -1,3 +1,3 @@
 """Version information."""
 
-__version__ = "1.16.7"
+__version__ = "1.17.0"

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/cli/commands/build.py

@@ -30,6 +30,7 @@ from ..framework import BuildCommand
 @click.option("--keep-output", is_flag=True, help="Preserve existing output directory (default: clear before build)")
 @click.option("--docx", is_flag=True, help="Also export to DOCX format for collaborative review")
 @click.option("--resolve-dois", "-r", is_flag=True, help="Attempt to resolve missing DOIs (when using --docx)")
+@click.option("--split-si", is_flag=True, help="Split PDF into main and SI sections (__main.pdf and __si.pdf)")
 @click.option("--verbose", "-v", is_flag=True, help="Enable verbose output")
 @click.option("--quiet", "-q", is_flag=True, help="Suppress non-essential output")
 @click.option("--debug", "-d", is_flag=True, help="Enable debug output")
@@ -49,6 +50,7 @@ def build(
     keep_output: bool,
     docx: bool,
     resolve_dois: bool,
+    split_si: bool,
     verbose: bool,
     quiet: bool,
     debug: bool,
@@ -81,6 +83,10 @@ def build(
 
         $ rxiv pdf --docx --resolve-dois
 
+    **Split PDF into main and SI sections:**
+
+        $ rxiv pdf --split-si
+
     **Force regenerate all figures:**
 
         $ rxiv pdf --force-figures
@@ -108,6 +114,7 @@ def build(
         keep_output=keep_output,
         docx=docx,
         resolve_dois=resolve_dois,
+        split_si=split_si,
         debug=debug or verbose,
         quiet=quiet,
         container_mode=container_mode,

rxiv_maker-1.17.0/src/rxiv_maker/cli/commands/docx.py

@@ -0,0 +1,177 @@
+"""DOCX export command for rxiv-maker CLI."""
+
+import platform
+import shutil
+import subprocess
+
+import rich_click as click
+from rich.console import Console
+
+from ...core.logging_config import get_logger
+from ...core.managers.dependency_manager import DependencyStatus, get_dependency_manager
+from ...exporters.docx_exporter import DocxExporter
+
+logger = get_logger()
+console = Console()
+
+
+def _check_and_offer_poppler_installation(console: Console, quiet: bool, verbose: bool) -> None:
+    """Check poppler availability and offer automatic installation via brew.
+
+    Args:
+        console: Rich console for output
+        quiet: Whether to suppress output
+        verbose: Whether verbose mode is enabled
+    """
+    # Check if poppler is installed
+    manager = get_dependency_manager()
+    result = manager.check_dependency("pdftoppm")
+
+    if result.status == DependencyStatus.AVAILABLE:
+        if verbose:
+            console.print("[dim]✓ Poppler utilities available[/dim]")
+        return
+
+    # Poppler is missing - offer to install
+    system = platform.system()
+
+    if system == "Darwin" and shutil.which("brew"):
+        # macOS with Homebrew
+        if not quiet:
+            console.print("[yellow]⚠️  Poppler not found[/yellow]")
+            console.print("   Poppler is needed to embed PDF figures in DOCX files.")
+            console.print("   Without it, PDF figures will appear as placeholders.")
+            console.print()
+
+        if click.confirm("   Would you like to install poppler now via Homebrew?", default=True):
+            console.print("[cyan]Installing poppler...[/cyan]")
+            try:
+                result = subprocess.run(
+                    ["brew", "install", "poppler"],
+                    capture_output=True,
+                    text=True,
+                    timeout=300,
+                )
+
+                if result.returncode == 0:
+                    console.print("[green]✅ Poppler installed successfully![/green]")
+                    # Clear dependency cache so it gets re-checked
+                    manager.clear_cache()
+                else:
+                    console.print(f"[red]❌ Installation failed:[/red] {result.stderr}")
+                    console.print("   You can install manually with: brew install poppler")
+            except subprocess.TimeoutExpired:
+                console.print("[red]❌ Installation timed out[/red]")
+            except Exception as e:
+                console.print(f"[red]❌ Installation error:[/red] {e}")
+        else:
+            console.print("   [dim]Skipping poppler installation. PDF figures will show as placeholders.[/dim]")
+
+    elif system == "Linux":
+        # Linux
+        if not quiet:
+            console.print("[yellow]⚠️  Poppler not found[/yellow]")
+            console.print("   Install with: sudo apt install poppler-utils")
+            console.print()
+    else:
+        # Other platforms or brew not available
+        if not quiet:
+            console.print("[yellow]⚠️  Poppler not found[/yellow]")
+            console.print(f"   Install instructions: {result.resolution_hint}")
+            console.print()
+
+
+@click.command(context_settings={"help_option_names": ["-h", "--help"]})
+@click.argument(
+    "manuscript_path",
+    type=click.Path(exists=True, file_okay=False),
+    required=False,
+    metavar="[MANUSCRIPT_PATH]",
+)
+@click.option(
+    "--resolve-dois",
+    "-r",
+    is_flag=True,
+    help="Attempt to resolve missing DOIs from metadata",
+)
+@click.option(
+    "--no-footnotes",
+    is_flag=True,
+    help="Disable references section (citations only)",
+)
+@click.option("--verbose", "-v", is_flag=True, help="Enable verbose output")
+@click.option("--quiet", "-q", is_flag=True, help="Suppress non-essential output")
+def docx(
+    manuscript_path: str | None,
+    resolve_dois: bool,
+    no_footnotes: bool,
+    verbose: bool,
+    quiet: bool,
+) -> None:
+    """Export manuscript to DOCX format for collaborative review.
+
+    Generates a Word document with numbered citations, embedded figures,
+    and a complete references section with DOI links. Output is automatically
+    saved to the manuscript directory with the same naming pattern as the PDF.
+
+    **MANUSCRIPT_PATH**: Directory containing manuscript files.
+    Defaults to MANUSCRIPT/
+
+    **Output**: Automatically saved to MANUSCRIPT/YEAR__lastname_et_al__rxiv.docx
+
+    ## Examples
+
+    **Basic export:**
+
+        $ rxiv docx
+
+    **Export from custom directory:**
+
+        $ rxiv docx MY_PAPER/
+
+    **With DOI resolution for missing entries:**
+
+        $ rxiv docx --resolve-dois
+
+    **Without references section (citations only):**
+
+        $ rxiv docx --no-footnotes
+    """
+    try:
+        # Configure logging
+        if verbose:
+            logger.set_level("DEBUG")
+        elif quiet:
+            logger.set_level("WARNING")
+
+        # Set manuscript path
+        manuscript_path = manuscript_path or "MANUSCRIPT"
+
+        # Create exporter
+        if not quiet:
+            console.print("[cyan]📄 Exporting manuscript to DOCX...[/cyan]")
+
+        exporter = DocxExporter(
+            manuscript_path=manuscript_path,
+            resolve_dois=resolve_dois,
+            include_footnotes=not no_footnotes,
+        )
+
+        # Pre-flight check for poppler (if manuscript contains PDF figures)
+        _check_and_offer_poppler_installation(console, quiet, verbose)
+
+        # Perform export
+        docx_path = exporter.export()
+
+        # Success message
+        if not quiet:
+            console.print(f"[green]✅ DOCX exported:[/green] {docx_path}")
+
+    except FileNotFoundError as e:
+        console.print(f"[red]❌ Error:[/red] {e}")
+        raise click.Abort() from e
+    except Exception as e:
+        if verbose:
+            logger.error(f"DOCX export failed: {e}")
+        console.print(f"[red]❌ Export failed:[/red] {e}")
+        raise click.Abort() from e

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/cli/framework/workflow_commands.py

@@ -142,6 +142,7 @@ class BuildCommand(BaseCommand):
         keep_output: bool = False,
         docx: bool = False,
         resolve_dois: bool = False,
+        split_si: bool = False,
         debug: bool = False,
         quiet: bool = False,
         container_mode: Optional[str] = None,
@@ -156,6 +157,7 @@ class BuildCommand(BaseCommand):
             keep_output: Preserve existing output directory
             docx: Also export to DOCX format
             resolve_dois: Attempt to resolve missing DOIs (for DOCX export)
+            split_si: Split PDF into main and SI sections
             debug: Enable debug output
             quiet: Suppress non-critical warnings
             container_mode: Container behavior mode
@@ -223,6 +225,10 @@ class BuildCommand(BaseCommand):
             if docx:
                 self._export_docx(resolve_dois=resolve_dois, quiet=quiet, debug=debug)
 
+            # Split PDF if requested
+            if split_si:
+                self._split_pdf(pdf_path, quiet=quiet, debug=debug)
+
             # Show helpful tips after successful build
             self._show_build_tips()
 
@@ -252,11 +258,69 @@ class BuildCommand(BaseCommand):
                 self.console.print(f"[green]✅ DOCX exported:[/green] {docx_path}")
 
         except Exception as e:
-            self.console.print(f"[yellow]⚠️  DOCX export failed:[/yellow] {e}", err=True)
+            self.console.print(f"[yellow]⚠️  DOCX export failed:[/yellow] {e}")
+            if debug:
+                import traceback
+
+                self.console.print(f"[dim]{traceback.format_exc()}[/dim]")
+
+    def _split_pdf(self, pdf_path: Path, quiet: bool = False, debug: bool = False) -> None:
+        """Split PDF into main and SI sections after successful PDF build.
+
+        Args:
+            pdf_path: Path to the generated PDF
+            quiet: Suppress non-essential output
+            debug: Enable debug output
+        """
+        try:
+            from ...processors.yaml_processor import extract_yaml_metadata
+            from ...utils.file_helpers import find_manuscript_md
+            from ...utils.pdf_splitter import split_pdf
+            from ...utils.pdf_utils import get_custom_pdf_filename
+
+            if not quiet:
+                self.console.print("\n[cyan]✂️  Splitting PDF into main and SI sections...[/cyan]")
+
+            # Split the PDF
+            main_path, si_path = split_pdf(pdf_path)
+
+            if main_path and si_path:
+                # Extract metadata to generate custom filename
+                manuscript_md = find_manuscript_md(str(self.path_manager.manuscript_path))
+                yaml_metadata = extract_yaml_metadata(str(manuscript_md))
+
+                # Get base filename (e.g., "2025__saraiva_et_al__rxiv.pdf")
+                base_filename = get_custom_pdf_filename(yaml_metadata)
+                base_name = base_filename.replace(".pdf", "")
+
+                # Generate final filenames with __main and __si suffixes
+                main_filename = f"{base_name}__main.pdf"
+                si_filename = f"{base_name}__si.pdf"
+
+                # Copy split files to manuscript directory
+                final_main_path = self.path_manager.manuscript_path / main_filename
+                final_si_path = self.path_manager.manuscript_path / si_filename
+
+                shutil.copy2(main_path, final_main_path)
+                shutil.copy2(si_path, final_si_path)
+
+                if not quiet:
+                    self.console.print("[green]✅ PDF split successfully:[/green]")
+                    self.console.print(f"   📄 Main: {final_main_path}")
+                    self.console.print(f"   📄 SI: {final_si_path}")
+            elif main_path is None and si_path is None:
+                if not quiet:
+                    self.console.print("[yellow]⚠️  Could not split PDF: SI section marker not found[/yellow]")
+            else:
+                if not quiet:
+                    self.console.print("[yellow]⚠️  PDF splitting partially failed[/yellow]")
+
+        except Exception as e:
+            self.console.print(f"[yellow]⚠️  PDF splitting failed:[/yellow] {e}")
             if debug:
                 import traceback
 
-                self.console.print(f"[dim]{traceback.format_exc()}[/dim]", err=True)
+                self.console.print(f"[dim]{traceback.format_exc()}[/dim]")
 
     def _show_build_tips(self) -> None:
         """Show helpful tips after successful PDF build."""

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/converters/citation_processor.py

@@ -202,9 +202,11 @@ def extract_citations_from_text(text: MarkdownContent) -> list[CitationKey]:
         backtick_patterns.append(match.group(0))
         return f"__BACKTICK_PATTERN_{len(backtick_patterns) - 1}__"
 
-    # Match both single backticks `...` and triple backticks ```...```
-    text_cleaned = re.sub(r"`[^`]+`", protect_backticks, text)
-    text_cleaned = re.sub(r"```.*?```", protect_backticks, text_cleaned, flags=re.DOTALL)
+    # IMPORTANT: Match triple backticks FIRST, then single backticks
+    # This prevents the single-backtick pattern from matching across triple-backtick blocks
+    # (e.g., from a ` before ```latex to the first ` inside the code block)
+    text_cleaned = re.sub(r"```.*?```", protect_backticks, text, flags=re.DOTALL)
+    text_cleaned = re.sub(r"`[^`]+`", protect_backticks, text_cleaned)
 
     # Find bracketed multiple citations
     bracketed_matches = re.findall(r"\[(@[^]]+)\]", text_cleaned)

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/core/managers/config_manager.py

@@ -343,6 +343,7 @@ class ConfigManager:
             "bibliography": {"file": "03_REFERENCES.bib", "style": "nature"},
             "citation_style": "numbered",
             "enable_inline_doi_resolution": False,
+            "docx": {"hide_si": False, "figures_at_end": False},
             "cache": {"enabled": True, "ttl_hours": 24},
             "version": "1.0",
             "acknowledge_rxiv_maker": True,

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/core/managers/dependency_manager.py

@@ -487,6 +487,18 @@ class DependencyManager:
                 )
             )
 
+        # Poppler utilities for PDF to image conversion (DOCX export)
+        self.register_dependency(
+            DependencySpec(
+                name="pdftoppm",
+                type=DependencyType.SYSTEM_BINARY,
+                required=False,  # Only needed for DOCX export with PDF figures
+                alternatives=["pdfinfo"],
+                contexts={"docx", "export"},
+                install_hint="macOS: brew install poppler | Linux: sudo apt install poppler-utils",
+            )
+        )
+
     def register_dependency(self, spec: DependencySpec) -> None:
         """Register a dependency specification.
 

{rxiv_maker-1.16.7 → rxiv_maker-1.17.0}/src/rxiv_maker/exporters/docx_citation_mapper.py

@@ -13,6 +13,102 @@ from ..converters.citation_processor import extract_citations_from_text
 class CitationMapper:
     """Maps citation keys to sequential numbers for DOCX export."""
 
+    @staticmethod
+    def _format_citation_ranges(text: str) -> str:
+        """Format consecutive citations as ranges.
+
+        Converts patterns like [1][2][3] to [1-3], [15][16] to [15-16], etc.
+        Also formats comma-separated lists like [1, 2, 3] to [1-3].
+
+        Args:
+            text: Text with numbered citations
+
+        Returns:
+            Text with consecutive citations formatted as ranges
+
+        Example:
+            >>> CitationMapper._format_citation_ranges("text [1][2][3] more")
+            'text [1-3] more'
+            >>> CitationMapper._format_citation_ranges("text [1, 2, 3] more")
+            'text [1-3] more'
+            >>> CitationMapper._format_citation_ranges("text [1][3][4] more")
+            'text [1][3-4] more'
+        """
+
+        # Pattern 1: Handle adjacent bracketed citations [1][2][3] or [1] [2] [3]
+        def combine_adjacent(match_obj):
+            # Extract all numbers from consecutive brackets (allowing spaces between)
+            numbers = [int(n) for n in re.findall(r"\[(\d+)\]", match_obj.group(0))]
+            return CitationMapper._format_number_list(numbers)
+
+        # Find sequences of adjacent bracketed numbers (with optional spaces between)
+        text = re.sub(r"(?:\[\d+\]\s*){2,}", combine_adjacent, text)
+
+        # Pattern 2: Handle comma-separated citations within single brackets [1, 2, 3]
+        def combine_comma_separated(match_obj):
+            # Extract all numbers from comma-separated list
+            numbers_str = match_obj.group(1)
+            numbers = [int(n.strip()) for n in numbers_str.split(",")]
+            return CitationMapper._format_number_list(numbers)
+
+        text = re.sub(r"\[([\d,\s]+)\]", combine_comma_separated, text)
+
+        return text
+
+    @staticmethod
+    def _format_number_list(numbers: List[int]) -> str:
+        """Format a list of citation numbers as ranges.
+
+        Args:
+            numbers: List of citation numbers
+
+        Returns:
+            Formatted string with ranges
+
+        Example:
+            >>> CitationMapper._format_number_list([1, 2, 3, 5, 6, 8])
+            '[1-3, 5-6, 8]'
+            >>> CitationMapper._format_number_list([15, 16])
+            '[15-16]'
+            >>> CitationMapper._format_number_list([1, 3, 5])
+            '[1, 3, 5]'
+        """
+        if not numbers:
+            return "[]"
+
+        # Sort numbers
+        sorted_nums = sorted(set(numbers))
+
+        # Build ranges
+        ranges = []
+        start = sorted_nums[0]
+        end = sorted_nums[0]
+
+        for num in sorted_nums[1:]:
+            if num == end + 1:
+                # Continue current range
+                end = num
+            else:
+                # End current range and start new one
+                if start == end:
+                    # Single number
+                    ranges.append(str(start))
+                else:
+                    # Range (including 2 consecutive numbers like 15-16)
+                    ranges.append(f"{start}-{end}")
+                start = num
+                end = num
+
+        # Add final range
+        if start == end:
+            # Single number
+            ranges.append(str(start))
+        else:
+            # Range (including 2 consecutive numbers like 15-16)
+            ranges.append(f"{start}-{end}")
+
+        return f"[{', '.join(ranges)}]"
+
     def create_mapping(self, citations: List[str]) -> Dict[str, int]:
         """Create citation key → number mapping.
 
@@ -121,4 +217,7 @@ class CitationMapper:
         for i, pattern in enumerate(email_patterns):
             text = text.replace(f"__EMAIL_PATTERN_{i}__", pattern)
 
+        # Format consecutive citations as ranges (e.g., [1][2][3] -> [1-3])
+        text = self._format_citation_ranges(text)
+
         return text