Merge pull request #36 from pmaher86/weasyprint

move to weasyprint
pmaher86 · Oct 7, 2024 · 31884d0 · 31884d0
2 parents cb3961f + 22ac3dd
commit 31884d0
Show file tree

Hide file tree

Showing 6 changed files with 168 additions and 77 deletions.
diff --git a/.github/workflows/build-and-test.yaml b/.github/workflows/build-and-test.yaml
@@ -22,13 +22,8 @@ jobs:
           pip install --upgrade pip
           pip install -r requirements.txt
 
-      - name: Install wkhtmltopdf
-        run: |
-          sudo apt-get update
-          sudo apt-get -y install wkhtmltopdf
-
       - name: Install package
-        run: pip install .[dev]
+        run: pip install .[dev,pdf]
 
       - name: Run pre-commit
         run: pre-commit run --show-diff-on-failure --color=always --all-files

diff --git a/README.md b/README.md
@@ -1,7 +1,7 @@
 # Blacksquare
 ![Build Status](https://github.com/pmaher86/blacksquare/actions/workflows/build-and-test.yaml/badge.svg) ![Documentation Status](https://readthedocs.org/projects/blacksquare/badge/?version=latest)
 
-Blacksquare is a Python package for crossword creators. It aims to be an intuitive interface for working with crossword puzzles programmatically. It also has tools for finding valid fills, and HTML rendering that plugs nicely into Jupyter notebooks. Blacksquare supports import and export from the .puz format via [puzpy](https://github.com/alexdej/puzpy), as well as .pdf export in the [New York Times submission format](https://www.nytimes.com/puzzles/submissions/crossword) (requires [wkhtmltopdf](https://wkhtmltopdf.org/)).
+Blacksquare is a Python package for crossword creators. It aims to be an intuitive interface for working with crossword puzzles programmatically. It also has tools for finding valid fills, and HTML rendering that plugs nicely into Jupyter notebooks. Blacksquare supports import and export from the .puz format via [puzpy](https://github.com/alexdej/puzpy), as well as .pdf export in the [New York Times submission format](https://www.nytimes.com/puzzles/submissions/crossword) (requires the [pdf] extra).
 
 ## Native HTML rendering in Jupyter
 ![Jupyter example](assets/jupyter.png?raw=true)
@@ -173,7 +173,9 @@ There's clearly some extra curation that could be done to improve the word list,
 ## Installation
 `pip install blacksquare`
 
-You'll also need to install [wkhtmltopdf](https://wkhtmltopdf.org/) for .pdf export to work.
+or if you want to enable pdf export
+
+`pip install blacksquare[pdf]`
 
 ## Future plans
 Blacksquare is not a GUI application and isn't intended to be one. Blacksquare is also not a package for solving crossword puzzles.

diff --git a/pyproject.toml b/pyproject.toml
@@ -6,15 +6,7 @@ name = "blacksquare"
 description = "A package for creating crossword puzzles"
 authors = [{ name = "Patrick Maher", email = "pmaher86@gmail.com" }]
 readme = "README.md"
-dependencies = [
-    "networkx",
-    "numpy",
-    "pandas",
-    "pdfkit",
-    "puzpy",
-    "pypdf",
-    "rich",
-]
+dependencies = ["networkx", "numpy", "pandas", "puzpy", "rich"]
 classifiers = [
     "Programming Language :: Python :: 3",
     "License :: OSI Approved :: MIT License",
@@ -28,6 +20,7 @@ Source = "https://github.com/pmaher86/blacksquare"
 
 [project.optional-dependencies]
 dev = ["pre-commit", "pytest", "uv"]
+pdf = ["pypdf", "weasyprint"]
 
 [tool.setuptools_scm]
 local_scheme = "no-local-version"

diff --git a/requirements.txt b/requirements.txt
@@ -1,11 +1,21 @@
 # This file was autogenerated by uv via the following command:
-#    uv pip compile --extra=dev pyproject.toml -o requirements.txt
+#    uv pip compile --extra dev --extra pdf pyproject.toml -o requirements.txt
+brotli==1.1.0
+    # via fonttools
+cffi==1.17.1
+    # via weasyprint
 cfgv==3.4.0
     # via pre-commit
+cssselect2==0.7.0
+    # via weasyprint
 distlib==0.3.8
     # via virtualenv
 filelock==3.16.1
     # via virtualenv
+fonttools==4.54.1
+    # via weasyprint
+html5lib==1.1
+    # via weasyprint
 identify==2.6.1
     # via pre-commit
 iniconfig==2.0.0
@@ -26,8 +36,8 @@ packaging==23.1
     # via pytest
 pandas==2.1.1
     # via blacksquare (pyproject.toml)
-pdfkit==1.0.0
-    # via blacksquare (pyproject.toml)
+pillow==10.4.0
+    # via weasyprint
 platformdirs==4.3.6
     # via virtualenv
 pluggy==1.3.0
@@ -36,10 +46,16 @@ pre-commit==3.8.0
     # via blacksquare (pyproject.toml)
 puzpy==0.2.5
     # via blacksquare (pyproject.toml)
+pycparser==2.22
+    # via cffi
+pydyf==0.11.0
+    # via weasyprint
 pygments==2.16.1
     # via rich
 pypdf==3.16.2
     # via blacksquare (pyproject.toml)
+pyphen==0.16.0
+    # via weasyprint
 pytest==7.4.2
     # via blacksquare (pyproject.toml)
 python-dateutil==2.8.2
@@ -51,10 +67,25 @@ pyyaml==6.0.2
 rich==13.5.3
     # via blacksquare (pyproject.toml)
 six==1.16.0
-    # via python-dateutil
+    # via
+    #   html5lib
+    #   python-dateutil
+tinycss2==1.3.0
+    # via
+    #   cssselect2
+    #   weasyprint
 tzdata==2023.3
     # via pandas
 uv==0.4.18
     # via blacksquare (pyproject.toml)
 virtualenv==20.26.6
     # via pre-commit
+weasyprint==62.3
+    # via blacksquare (pyproject.toml)
+webencodings==0.5.1
+    # via
+    #   cssselect2
+    #   html5lib
+    #   tinycss2
+zopfli==0.2.3
+    # via fonttools
diff --git a/src/blacksquare/crossword.py b/src/blacksquare/crossword.py
@@ -8,15 +8,14 @@
 
 import networkx as nx
 import numpy as np
-import pdfkit
 import puz
 import rich.box
-from pypdf import PdfReader, PdfWriter
 from rich.console import Console
 from rich.live import Live
 from rich.table import Table
 
 from blacksquare.cell import Cell
+from blacksquare.html import CSS_TEMPLATE
 from blacksquare.symmetry import Symmetry
 from blacksquare.types import (
     CellIndex,
@@ -29,6 +28,13 @@
 from blacksquare.word import Word
 from blacksquare.word_list import DEFAULT_WORDLIST, WordList
 
+try:
+    import pypdf
+    import weasyprint
+except ImportError:
+    weasyprint = None
+    pypdf = None
+
 BLACK, EMPTY = SpecialCellValue.BLACK, SpecialCellValue.EMPTY
 ACROSS, DOWN = Direction.ACROSS, Direction.DOWN
 
@@ -203,15 +209,36 @@ def to_pdf(
                 name, address, etc.). Each list element will be one line in the header.
         """
 
+        if weasyprint is None:
+            raise ImportError(
+                "Can't import weasyprint, run pip install blacksquare[pdf] to install."
+            )
+
         header_html = "<br />".join(header) if header else ""
         grid_html = f"""
             <html>
-            <head><meta charset="utf-8"></head>
+            <head><meta charset="utf-8">
+            <style>
+            @page {{
+                margin:0.25 in;
+                margin-bottom: 0;
+            }}
+
+            @media print {{
+            div {{
+                break-inside: avoid-page !important;
+            }}
+            }}
+
+
+            </style>
+            </head>
             <body>
-            <div style='font-size:18pt;'>
+            <div style='font-size:14pt; break-after: avoid-page !important;'>
                 {header_html}
             </div>
-            <div style='position:absolute;left:50%;top:50%;transform: translate(-50%, -50%);'>
+            <br /> <br /> <br /> <br />
+            <div style='margin: auto;'>
                 {self._grid_html(size_px=600)}
             </div>
             </body></html>
@@ -235,7 +262,7 @@ def clue_rows(direction):
                     table {{
                         text-align:left;
                         width:100%;
-                        font-size:18pt;
+                        font-size:16pt;
                         border-spacing:1rem;
                     }}
                 </style>
@@ -250,21 +277,10 @@ def clue_rows(direction):
             </tbody></table>
             </body></html>
         """
-        merger = PdfWriter()
+        merger = pypdf.PdfWriter()
         for html_page in [grid_html, clue_html]:
-            pdf = pdfkit.from_string(
-                html_page,
-                False,
-                options={
-                    "quiet": None,
-                    "margin-top": "0.5in",
-                    "margin-right": "0.5in",
-                    "margin-bottom": "0.5in",
-                    "margin-left": "0.5in",
-                    "encoding": "UTF-8",
-                },
-            )
-            merger.append(PdfReader(io.BytesIO(pdf)))
+            pdf = weasyprint.HTML(string=html_page, encoding="UTF-8").write_pdf()
+            merger.append(pypdf.PdfReader(io.BytesIO(pdf)))
         merger.write(str(filename))
         merger.close()
 
@@ -761,45 +777,45 @@ def _grid_html(self, size_px: Optional[int] = None) -> str:
         size_px = size_px or self.display_size_px
         # Random suffix is a hack to ensure correct display in Jupyter settings
         suffix = token_hex(4)
-        circle_string = f"<div class='circle{suffix}'> </div>"
-        row_elems = []
-        for r in range(self._num_rows):
-            cells = []
-            for c in range(self._num_cols):
-                number = self._numbers[r, c]
-                cells.append(
-                    f"""<td class='xw{suffix}{f" black{suffix}" if self._grid[r, c] == BLACK else ""}{f" gray{suffix}" if self._grid[r, c].shaded else ""}'>
-                        <div class='number{suffix}'> {number if number else ""}</div>
-                        <div class='value{suffix}'>
-                            {self._grid[r,c].str if self._grid[r,c] != BLACK else ""}
-                        </div>
-                        {circle_string if self._grid[r,c].circled else ""}
-                    </td >"""
-                )
-            row_elems.append(f"<tr class='xw{suffix}'>{''.join(cells)}</tr>")
+        cells = []
+        for c in self.itercells():
+            c.number
+            cell_number_span = f'<span class="cell-number">{c.number or ""}</span>'
+            letter_span = f'<span class="letter">{c.value if c!=BLACK else ""}</span>'
+            circle_span = '<span class="circle"></span>'
+            if c == BLACK:
+                extra_class = " black"
+            elif c.shaded:
+                extra_class = " gray"
+            else:
+                extra_class = ""
+            cell_div = f"""
+            <div class="crossword-cell{suffix}{extra_class}">
+                {cell_number_span}
+                {letter_span}
+                {circle_span if c.circled else ""}
+            </div>
+            """
+            cells.append(cell_div)
         aspect_ratio = self.num_rows / self.num_cols
         cell_size = size_px / max(self.num_rows, self.num_cols)
-        return """
-        <div>
-        <style scoped>
-        table.xw{suffix} {{table-layout:fixed; background-color:white;width:{width}px;height:{height}px;}}
-        td.xw{suffix} {{outline: 2px solid black;outline-offset: -1px;position: relative;font-family: Arial, Helvetica, sans-serif;}}
-        tr.xw{suffix} {{background-color: white !important;}}
-        .number{suffix} {{position: absolute;top: 2px;left: 2px;font-size: {num_font}px;font-weight: normal;user-select: none; color: black;}}
-        .value{suffix} {{position: absolute;bottom:0;left: 50%;font-weight: bold;font-size: {val_font}px; transform: translate(-50%, 0%); color: black;}}
-        .black{suffix} {{background-color: black;}}
-        .gray{suffix} {{background-color: lightgrey;}}
-        .circle{suffix} {{position: absolute; border-radius: 50%; border: 1px solid black; right: 0px; left: 0px; top: 0px; bottom: 0px;}}
-        </style>
-        <table class='xw{suffix}'><tbody>
-            {rows}
-        </tbody</table>
-        </div>
-        """.format(
+        css = CSS_TEMPLATE.format(
+            num_cols=self.num_cols,
             height=size_px * min(1, aspect_ratio),
             width=size_px * min(1, 1 / aspect_ratio),
-            rows="\n".join(row_elems),
+            num_font_size=int(cell_size * 0.3),
+            val_font_size=int(cell_size * 0.6),
+            circle_dim=cell_size - 1,
             suffix=suffix,
-            num_font=int(cell_size * 0.3),
-            val_font=int(cell_size * 0.6),
         )
+        cells_html = "\n".join(cells)
+        return f"""
+        <div>
+            <style scoped>
+                {css}
+            </style>
+            <div class="crossword{suffix}">
+                {cells_html}
+            </div>
+        </div>
+        """
diff --git a/src/blacksquare/html.py b/src/blacksquare/html.py
@@ -0,0 +1,54 @@
+CSS_TEMPLATE = """
+.crossword{suffix} {{
+    display: grid;
+    grid-template-columns: repeat({num_cols}, 1fr);
+    grid-auto-rows: 1fr;
+    gap: 0px;
+    width: {width}px;
+    height: {height}px;
+}}
+
+.crossword-cell{suffix} {{
+    position: relative;
+    background-color: white;
+    outline: 1px solid black;
+    font-family: Arial, Helvetica, sans-serif;
+    aspect-ratio: 1;
+}}
+
+.crossword-cell{suffix} .cell-number {{
+    position: absolute;
+    top: 2px;
+    left: 2px;
+    font-size: {num_font_size}px;
+    color: black;
+    user-select: none;
+}}
+
+.crossword-cell{suffix} .letter {{
+    font-size: {val_font_size}px;
+    position: absolute;
+    bottom: 0;
+    left: 50%;
+    transform: translate(-50%, 0%);
+    color: black;
+}}
+
+.black {{
+    background-color: black;
+    outline: 1px solid gray;
+}}
+
+.gray {{
+    background-color: lightgray;
+}}
+
+.crossword-cell{suffix} .circle {{
+    position: absolute;
+    border-radius: 50%;
+    border: 1px solid black;
+    height: {circle_dim}px;
+    width: {circle_dim}px;
+    margin: -1px;
+}}
+"""