Better handling of text overflow in FPDF.write() & FPDF.write_html() - …

…fix #847 (#850)

CHANGELOG.md

@@ -27,6 +27,7 @@ This can also be enabled programmatically with `warnings.simplefilter('default',
 ### Fixed
 - [`FPDF.table()`](https://pyfpdf.github.io/fpdf2/fpdf/fpdf.html#fpdf.fpdf.FPDF.table): the `colspan` setting has been fixed - [documentation](https://pyfpdf.github.io/fpdf2/Tables.html#column-span)
 - [`FPDF.image()`](https://pyfpdf.github.io/fpdf2/fpdf/fpdf.html#fpdf.fpdf.FPDF.image): allowing images path starting with `data` to be passed as input
+- text overflow is better handled by `FPDF.write()` & `FPDF.write_html()` - _cf._ [issue #847](https://github.com/PyFPDF/fpdf2/issues/847)
 - the initial text color is preserved when using `FPDF.write_html()` - _cf._ [issue #846](https://github.com/PyFPDF/fpdf2/issues/846)
 ### Deprecated
 - the `center` optional parameter of [`FPDF.cell()`](https://pyfpdf.github.io/fpdf2/fpdf/fpdf.html#fpdf.fpdf.FPDF.cell) is **no more** deprecated, as it allows for horizontal positioning, which is different from text alignment control with `align="C"`

fpdf/fonts.py

@@ -61,6 +61,9 @@ def __init__(self, fpdf, fontkey, style):
         self.fontkey = fontkey
         self.emphasis = TextEmphasis.coerce(style)
 
+    def __repr__(self):
+        return f"CoreFont(i={self.i}, fontkey={self.fontkey})"
+
 
 class TTFFont:
     __slots__ = (

fpdf/fpdf.py

@@ -3605,7 +3605,7 @@ def write(
         # first line from current x position to right margin
         first_width = self.w - self.x - self.r_margin
         txt_line = multi_line_break.get_line_of_given_width(
-            first_width - 2 * self.c_margin, wordsplit=False
+            first_width - 2 * self.c_margin,
         )
         # remaining lines fill between margins
         full_width = self.w - self.l_margin - self.r_margin

fpdf/line_break.py

@@ -38,12 +38,10 @@ def __init__(
         self.link = link
 
     def __repr__(self):
-        gstate = self.graphics_state.copy()
-        if "current_font" in gstate:
-            del gstate["current_font"]  # TMI
         return (
             f"Fragment(characters={self.characters},"
-            f" graphics_state={gstate}, k={self.k}, link={self.link})"
+            f" graphics_state={self.graphics_state},"
+            f" k={self.k}, link={self.link})"
         )
 
     @property
@@ -394,18 +392,14 @@ def __init__(
         self.idx_last_forced_break = None
 
     # pylint: disable=too-many-return-statements
-    def get_line_of_given_width(self, maximum_width: float, wordsplit: bool = True):
+    def get_line_of_given_width(self, maximum_width: float):
         first_char = True  # "Tw" ignores the first character in a text object.
         idx_last_forced_break = self.idx_last_forced_break
         self.idx_last_forced_break = None
 
         if self.fragment_index == len(self.styled_text_fragments):
             return None
 
-        last_fragment_index = self.fragment_index
-        last_character_index = self.character_index
-        line_full = False
-
         current_line = CurrentLine(print_sh=self.print_sh)
         while self.fragment_index < len(self.styled_text_fragments):
             current_fragment = self.styled_text_fragments[self.fragment_index]
@@ -442,9 +436,6 @@ def get_line_of_given_width(self, maximum_width: float, wordsplit: bool = True):
                     ) = current_line.automatic_break(self.justify)
                     self.character_index += 1
                     return line
-                if not wordsplit:
-                    line_full = True
-                    break
                 if idx_last_forced_break == self.character_index:
                     raise FPDFException(
                         "Not enough horizontal space to render a single character"
@@ -464,12 +455,6 @@ def get_line_of_given_width(self, maximum_width: float, wordsplit: bool = True):
 
             self.character_index += 1
 
-        if line_full and not wordsplit:
-            # roll back and return empty line to trigger continuation
-            # on the next line.
-            self.fragment_index = last_fragment_index
-            self.character_index = last_character_index
-            return CurrentLine().manual_break(self.justify)
         if current_line.width:
             return current_line.manual_break()
         return None

test/text/test_line_break.py

@@ -1129,3 +1129,13 @@ def test_trim_trailing_spaces():
     cl.fragments = [frag]
     res = cl.trim_trailing_spaces()
     assert res is None
+
+
+def test_line_break_no_initial_newline():  # issue-847
+    text = "X" * 50
+    alphabet = {"normal": {}}
+    alphabet["normal"]["X"] = 4.7
+    fragments = [FxFragment(alphabet, text, _gs_normal, 1)]
+    multi_line_break = MultiLineBreak(fragments)
+    text_line = multi_line_break.get_line_of_given_width(188)
+    assert text_line.fragments

test/text/test_unbreakable.py

@@ -167,12 +167,12 @@ def test_multi_cell_table_unbreakable_with_split_only(tmp_path):  # issue 359
 
     pdf.ln()
 
-    with pdf.unbreakable() as doc:
-        for _ in range(4):
-            for row in data:
-                max_no_of_lines_in_cell = 1
-                for cell in row:
-                    with pytest.warns(DeprecationWarning, match=expected_warn):
+    with pytest.warns(DeprecationWarning, match=expected_warn):
+        with pdf.unbreakable() as doc:
+            for _ in range(4):
+                for row in data:
+                    max_no_of_lines_in_cell = 1
+                    for cell in row:
                         result = doc.multi_cell(
                             cell_width,
                             l_height,
@@ -184,39 +184,39 @@ def test_multi_cell_table_unbreakable_with_split_only(tmp_path):  # issue 359
                             max_line_height=l_height,
                             split_only=True,
                         )
-                    no_of_lines_in_cell = len(result)
-                    if no_of_lines_in_cell > max_no_of_lines_in_cell:
-                        max_no_of_lines_in_cell = no_of_lines_in_cell
-                no_of_lines_list.append(max_no_of_lines_in_cell)
-
-            for j, row in enumerate(data):
-                cell_height = no_of_lines_list[j] * l_height
-                for cell in row:
-                    if j == 0:
-                        doc.multi_cell(
-                            cell_width,
-                            cell_height,
-                            "**" + cell + "**",
-                            border=1,
-                            fill=False,
-                            align="L",
-                            new_x="RIGHT",
-                            new_y="TOP",
-                            max_line_height=l_height,
-                            markdown=False,
-                        )
-                    else:
-                        doc.multi_cell(
-                            cell_width,
-                            cell_height,
-                            cell,
-                            border=1,
-                            align="L",
-                            new_x="RIGHT",
-                            new_y="TOP",
-                            max_line_height=l_height,
-                        )
-                doc.ln(cell_height)
+                        no_of_lines_in_cell = len(result)
+                        if no_of_lines_in_cell > max_no_of_lines_in_cell:
+                            max_no_of_lines_in_cell = no_of_lines_in_cell
+                    no_of_lines_list.append(max_no_of_lines_in_cell)
+
+                for j, row in enumerate(data):
+                    cell_height = no_of_lines_list[j] * l_height
+                    for cell in row:
+                        if j == 0:
+                            doc.multi_cell(
+                                cell_width,
+                                cell_height,
+                                "**" + cell + "**",
+                                border=1,
+                                fill=False,
+                                align="L",
+                                new_x="RIGHT",
+                                new_y="TOP",
+                                max_line_height=l_height,
+                                markdown=False,
+                            )
+                        else:
+                            doc.multi_cell(
+                                cell_width,
+                                cell_height,
+                                cell,
+                                border=1,
+                                align="L",
+                                new_x="RIGHT",
+                                new_y="TOP",
+                                max_line_height=l_height,
+                            )
+                    doc.ln(cell_height)
 
     assert_pdf_equal(
         pdf, HERE / "multi_cell_table_unbreakable_with_split_only.pdf", tmp_path

test/text/test_varied_fragments.py

@@ -29,7 +29,7 @@ def write_fragments(self, frags, align=Align.L):
         # first line from current x position to right margin
         first_width = self.w - self.x - self.r_margin
         text_line = multi_line_break.get_line_of_given_width(
-            first_width - 2 * self.c_margin, wordsplit=False
+            first_width - 2 * self.c_margin
         )
         # remaining lines fill between margins
         full_width = self.w - self.l_margin - self.r_margin

test/text/test_write.py

@@ -1,14 +1,14 @@
 from pathlib import Path
 
-import fpdf
+from fpdf import FPDF
 from test.conftest import assert_pdf_equal, LOREM_IPSUM
 
 HERE = Path(__file__).resolve().parent
 FONTS_DIR = HERE.parent / "fonts"
 
 
 def test_write_page_break(tmp_path):
-    doc = fpdf.FPDF()
+    doc = FPDF()
     doc.add_page()
     doc.set_font("helvetica", size=24)
     doc.y = 20
@@ -18,8 +18,15 @@ def test_write_page_break(tmp_path):
 
 
 def test_write_soft_hyphen(tmp_path):
+    """
+    The current behaviour is close to CSS word-break: break-all
+    cf. https://developer.mozilla.org/en-US/docs/Web/CSS/overflow-wrap#comparing_overflow-wrap_word-break_and_hyphens
+    We used to prefer a line break over a word split without regards to soft hyphens:
+    https://github.com/PyFPDF/fpdf2/blob/2.7.4/test/text/write_soft_hyphen.pdf
+    But that caused issue with write_html(), cf. issue #847
+    """
     s = "Donau\u00addamp\u00adfschiff\u00adfahrts\u00adgesellschafts\u00adkapitäns\u00admützen\u00adstreifen. "
-    doc = fpdf.FPDF()
+    doc = FPDF()
     doc.add_page()
     doc.set_font("helvetica", size=24)
     doc.y = 20
@@ -41,7 +48,7 @@ def test_write_soft_hyphen(tmp_path):
 
 def test_write_trailing_nl(tmp_path):  # issue #455
     """Each item in lines triggers a line break at the end."""
-    pdf = fpdf.FPDF()
+    pdf = FPDF()
     pdf.add_page()
     pdf.set_font("Times", size=16)
     lines = ["Hello\n", "Sweet\n", "World\n"]
@@ -53,7 +60,7 @@ def test_write_trailing_nl(tmp_path):  # issue #455
 
 def test_write_font_stretching(tmp_path):  # issue #478
     right_boundary = 60
-    pdf = fpdf.FPDF()
+    pdf = FPDF()
     pdf.add_page()
     # built-in font
     pdf.set_font("Helvetica", "", 8)
@@ -81,7 +88,7 @@ def test_write_font_stretching(tmp_path):  # issue #478
 
 
 def test_write_superscript(tmp_path):
-    pdf = fpdf.FPDF()
+    pdf = FPDF()
     pdf.add_page()
     pdf.set_font("Helvetica", "", 20)
 
@@ -131,7 +138,7 @@ def write_this():
 
 def test_write_char_wrap(tmp_path):  # issue #649
     right_boundary = 50
-    pdf = fpdf.FPDF()
+    pdf = FPDF()
     pdf.add_page()
     pdf.set_right_margin(pdf.w - right_boundary)
     pdf.set_font("Helvetica", "", 10)
@@ -150,3 +157,11 @@ def test_write_char_wrap(tmp_path):  # issue #649
     pdf.line(pdf.l_margin, 10, pdf.l_margin, 130)
     pdf.line(right_boundary, 10, right_boundary, 130)
     assert_pdf_equal(pdf, HERE / "write_char_wrap.pdf", tmp_path)
+
+
+def test_write_overflow_no_initial_newline(tmp_path):  # issue-847
+    pdf = FPDF()
+    pdf.add_page()
+    pdf.set_font(family="Helvetica", size=20)
+    pdf.write(7, "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX")
+    assert_pdf_equal(pdf, HERE / "write_overflow_no_initial_newline.pdf", tmp_path)