rawmaker 2.40.3__py3-none-any.whl

rawmaker/fonts/__init__.py

@@ -0,0 +1,8 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2020-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================

rawmaker/fonts/parser.py

@@ -0,0 +1,354 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2020-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================
+"""Font Parser
+
+PDF Font description:
+
+9.5. Introduction into Font Data Structures
+
+Font types
+
+    Type0
+    Type1           Type1
+                    MMType1: MultiMaster Font
+    Type3           stream of pdf graphic operators
+    TrueType        Based on TrueType font format
+    CIDFont         CIDFontType0
+                    CIDFontType2
+
+9.6.2.2 Standard Type 1 Fonts
+
+    uses compact encoding for glyph description and additional hints to print
+    on small sizes and solutions well.
+
+    PostScript 14 standard types:
+        Times-Roman, Helvectica, Courier, Symbol, Times-Bold,
+        Helvetica-Bold, Courier-Bold, ZapfDingbats, Times-Italic,
+        Helvetica-Oblique, Courier-Oblique, Times-BoldItalic,
+        Helvetica-BoldOblique, Courier-BoldOblique.
+
+9.6.2.3 MultiMasterFonts
+
+9.6.3. TrueTypeFonts
+
+9.6.4. Font Subsets
+
+    BaseFont
+    FontName
+
+    Tag(6 chars) +
+
+    Example: EOODIA+Poetica - name of a subset of Poetica, a Type 1 font.
+
+9.6.5 Type 3 Fonts
+
+    Defined by a stream of pdf graphic commands, no special support or hint
+    for very small characters.
+
+9.7.4 CIDFonts
+
+    CIDFont program contains glyph descriptions that are accessed using a CID
+    as a character selector.
+
+Summary:
+
+    Font Type 0
+        ('Helvetica - Bold', 16.70),
+        ('Times - Roman', 13.40),
+    Font Type 1, TrueType Fonts:
+        ('ZTJCPR + NimbusRomNo9L - MediItal', 11.60),
+        ('KCXMNX + TeX - feymr10', 10.70),
+"""
+
+import contextlib
+
+import iamraw
+import iamraw.fonts
+import serializeraw
+import utilo
+
+POSTSCRIPT_14_DEFAULT = {
+    'Courier',
+    'Courier-Bold',
+    'Courier-BoldOblique',
+    'Courier-Oblique',
+    'Helvectica',
+    'Helvetica-Bold',
+    'Helvetica-BoldOblique',
+    'Helvetica-Oblique',
+    'Symbol',
+    'Times-Bold',
+    'Times-BoldItalic',
+    'Times-Italic',
+    'Times-Roman',
+    'ZapfDingbats',
+}
+
+
+def font_fromraw(font: str, scale: float = 0.0, flags: int = 0) -> iamraw.Font:
+    """Parse `Font` from pdf representation, read the description above.
+
+    Args:
+        font(str): pdf standard font definition
+        scale(float): size of font(unit?)
+        flags(int): style of rendered font
+    Returns:
+        returns internal `Font` object with detected style and scale
+
+    >>> font_fromraw('Times-RomanRegularLight')
+    Font(pdfref='Times-RomanRegularLight',...)
+    >>> font_fromraw('CFSJNQ+LinBiolinumTB')
+    Font(pdfref='CFSJNQ+LinBiolinumTB',...weight=iamraw.fonts.Weight.BOLD,...)
+    """
+    pdfref = font
+    utilo.call('rawmaker.fonts.parser.font_fromraw')
+    utilo.debug('%s %.2f' % (str(font), scale))  # pylint:disable=C0209
+    flags = serializeraw.load_flags(flags)
+    # remove white spaces to avoid missing PostScript 14 language cause of
+    # containg white spaces, for example `Times - Roman` instead of
+    # `Times-Roman`.
+    font = font.replace(' ', '')
+    # parse different fonts
+    basefont = parse_basefont(font)
+    cidfont = parse_cidfont(font)
+    default = parse_default(font)
+    styled = parse_font_styled(font)
+    simple = parse_font_simple(font)
+    # select best font parsing
+    fontname, style = None, None
+    if cidfont is not None:
+        # cidfont at first, cause cidfont selector is the clearest and not
+        # ambigous.
+        fontname, style = cidfont
+    elif simple:
+        fontname, style = simple
+    elif styled:
+        fontname, style = styled
+    elif basefont is not None:
+        fontname, style = basefont
+    elif default is not None:
+        fontname, style = default
+    # use default style if no style is given
+    weight, style, stretch = style if style else (None, None, None)
+    # inform about parsing problem
+    if fontname is None or '+' in fontname or ',' in fontname:
+        utilo.error(f'detected fontname {fontname}; input: {font}')
+    font = iamraw.Font(
+        name=fontname,
+        scale=scale,
+        stretch=stretch,
+        style=style,
+        weight=weight,
+        flags=flags,
+        pdfref=pdfref,
+    )
+    return font
+
+
+def parse_basefont(font: str):
+    """\
+    >>> parse_basefont('Arial,Bold')
+    ('Arial',...Weight.BOLD...)
+    >>> parse_basefont('WTUVLZ+NimbusRomNo9L-Regu') is None
+    True
+    >>> parse_basefont('Times-RomanRegularLight')
+    ('Times-Roman', (...))
+    """
+    if '+' in font:
+        return None
+    # Example: Arial,Bold
+    fontname, raw_style = font, ''
+    with contextlib.suppress(ValueError):
+        fontname, raw_style = font.split(',')
+    style = parse_style(raw_style)
+    if not style:
+        return None
+    for name, *styles in STYLES:
+        if not set(style) & set(styles):
+            # nothing to replace
+            continue
+        fontname = utilo.rreplace(fontname, name, '')
+    return fontname, style
+
+
+def parse_cidfont(font: str):
+    """\
+    >>> parse_cidfont('CIDFont+F1')
+    ('F1', None)
+    """
+    cidfont = font.startswith('CIDFont+')
+    if not cidfont:
+        return None
+    # Example: CIDFont+F1
+    # remove cid tag and plus sign
+    fontname = font[8:]
+    return fontname, None
+
+
+def parse_postscript14(fontname: str):
+    if fontname not in POSTSCRIPT_14_DEFAULT:
+        return None
+    if '-' not in fontname:
+        # Courier
+        return fontname, None
+    # 'Courier-Oblique',
+    fontname, style = fontname.split('-')
+    style = parse_style(style)
+    return fontname, style
+
+
+def parse_default(font: str):
+    # Example: LGAZPG + SegoeUI, Bold
+    # remove base tag and plus sign
+    font = font[7:]
+    fontname, raw_style = font, ''
+
+    parsed = parse_postscript14(font)
+    if parsed:
+        # 'AIDZQU+Times-Roman' no style parsing is required
+        return parsed
+
+    style = None
+    with contextlib.suppress(ValueError):
+        fontname, raw_style = font.split(',')
+    with contextlib.suppress(ValueError):
+        fontname, raw_style = font.split('-')
+    style = parse_style(raw_style)
+    if not raw_style:
+        parsed = parse_font_simple(fontname)
+        if parsed:
+            fontname, style = parsed
+    if not style:
+        # TODO: FONT STYLE PARSER
+        return fontname, None
+    return fontname, style
+
+
+def parse_font_styled(font: str):
+    if not font.count('-') == 1 or '+' in font:
+        return None
+    name, style = font.split('-')
+    style = parse_style(style)
+
+    if not style:
+        return None
+    name = named(name)
+    return name, style
+
+
+def parse_font_simple(font: str):
+    if any(char in font for char in {'+', '-', ','}):  # pylint:disable=use-sequence-for-iteration
+        return None
+    styles = []
+    for item in STYLES:
+        if item[0] not in font:
+            continue
+        styles.append((item[1], item[2], item[3]))
+        font = font.replace(item[0], '')
+
+    weight, style, stretch = MEDIUM, NORMAL, REGULAR
+    for item in styles:
+        if item[0]:
+            weight = item[0]
+        if item[1]:
+            style = item[1]
+        if item[2]:
+            stretch = item[2]
+    return font, (weight, style, stretch)
+
+
+def named(font: str):
+    for item in STYLES:
+        font = font.replace(item[0], '')
+    return font
+
+
+def font_toraw(font: iamraw.Font) -> str:
+    """\
+    >>> font_toraw(iamraw.Font(name='Times-Roman', scale=8.95, weight=LIGHT, stretch=REGULAR))
+    'Times-RomanRegularLight'
+    """
+    result = font.name
+    selected = {font.weight, font.style, font.stretch}
+    if not any(selected):
+        # no style given, do not use default style
+        return f'CIDFont+{result}'
+    styles = [
+        ('Bd', BOLD, None, None),
+        ('Italic', None, ITALIC, None),
+        ('Medium', MEDIUM, None, None),
+        ('Oblique', None, OBLIQUE, None),
+        ('Regular', None, None, REGULAR),
+        ('Light', LIGHT, None, None),
+    ]
+    for raw, *items in styles:
+        if not any(item for item in items if item in selected):
+            continue
+        result += raw
+    return result
+
+
+def parse_style(raw_style):  # pylint:disable=R1260,R0912
+    save = raw_style
+    weight, style, stretch = LIGHT, NORMAL, REGULAR
+    for item in STYLES:
+        if item[0] not in raw_style:
+            continue
+        raw_style = raw_style.replace(item[0], '')
+        if item[1]:
+            weight = item[1]  # pylint:disable=R0204
+        if item[2]:
+            style = item[2]
+        if item[3]:
+            stretch = item[3]
+    if raw_style:  # TODO: Remove before going live
+        # at the end, everything must be replaced
+        utilo.error(f'unsupported style {save}, maybe a name: {raw_style}')
+    if raw_style:
+        return None
+    return weight, style, stretch
+
+
+BOLD = iamraw.Weight.BOLD
+ITALIC = iamraw.Style.ITALIC
+LIGHT = iamraw.Weight.LIGHT
+MEDIUM = iamraw.Weight.MEDIUM
+NORMAL = iamraw.Style.NORMAL
+OBLIQUE = iamraw.Style.OBLIQUE
+REGULAR = iamraw.Stretch.REGULAR
+
+# TODO: INVESTIGATE MT
+# BOLD = BOLD
+# MT = MEDIUM
+# HOW TO DEAL WITH BOLD MT?
+
+STYLES = [
+    ('Bd', BOLD, None, None),
+    ('Italic', None, ITALIC, None),
+    ('Ital', None, ITALIC, None),
+    # ('MI', MEDIUM, ITALIC, None),
+    ('Medium', MEDIUM, None, None),
+    ('Medi', MEDIUM, None, None),
+    # ('M', MEDIUM, None, None),  # HOW TO DEAL WITH ?BOLDMT?
+    ('MT', None, None, None),  # HOW TO DEAL WITH ?BOLDMT? TODO: DECIDE LATER
+    ('Bold', BOLD, None, None),  # HOW TO DEAL WITH ?BOLDMT?
+    ('Oblique', None, OBLIQUE, None),
+    ('Obli', None, OBLIQUE, None),
+    ('PSMT', MEDIUM, None, None),
+    ('PS', MEDIUM, None, None),
+    ('Regular', None, None, REGULAR),
+    ('Regu', None, None, REGULAR),
+    ('Rg', None, None, REGULAR),
+    ('Light', LIGHT, None, None),
+    ('TB', BOLD, None, None),
+]
+# LinBiolinumTB: TODO CHECK TB
+#  TODO: Roman converts TimesNewRoman to TimesNew. I could not verify if
+#  that is a smart necessary option or it will introduce more problems?
+# ('Roman', None, None, None),

rawmaker/images/__init__.py

@@ -0,0 +1,8 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2020-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================

rawmaker/images/info.py

@@ -0,0 +1,35 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2020-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================
+
+import iamraw
+import PIL.Image
+import utilo
+
+DEFAULT_DPI = (96, 96)
+
+
+def imageinfo(path: str, page: int, bounding: tuple) -> iamraw.ImageInformation:
+    assert isinstance(bounding, (iamraw.BoundingBox, tuple)), type(bounding)
+    try:
+        image = PIL.Image.open(path)
+        image.load()
+    except OSError as err:
+        utilo.error(err)
+        return None
+    width, height = image.size
+    # add default DPI to distinguish images and figures
+    dpi = image.info.get('dpi', DEFAULT_DPI)
+    result = iamraw.ImageInformation(
+        width=width,
+        height=height,
+        dpi=dpi,
+        page=page,
+        bounding=bounding,
+    )
+    return result

rawmaker/miner/__init__.py

@@ -0,0 +1,8 @@
+#==============================================================================
+# C O P Y R I G H T
+#------------------------------------------------------------------------------
+# Copyright (c) 2019-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+#==============================================================================

rawmaker/miner/char.py

@@ -0,0 +1,42 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2020-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================
+
+import iamraw
+
+import rawmaker.converter.basic
+import rawmaker.miner.text
+import rawmaker.parameter
+import rawmaker.patch.ltchar
+
+
+class CharPDFConvert(rawmaker.miner.text.PrecisePDFConverter):
+
+    def __init__(
+        self,
+        config: rawmaker.parameter.ParsingConfiguration = None,
+        imagewriter: callable = None,
+    ):
+        super().__init__()
+        self.laparams = None  # disable layout analysis
+
+    def receive_layout(self, ltpage):
+        rawmaker.converter.basic.FlippedLayoutAnalyzer.receive_layout(
+            self, ltpage)
+        chars = [
+            char for char in ltpage
+            if isinstance(char, rawmaker.patch.ltchar.PatchedLTChar)
+        ]
+        chars = sorted(chars, key=lambda x: x.bbox[0])  # x0 # pylint:disable=C3001
+        chars = sorted(chars, key=lambda x: x.bbox[3])  # y1 # pylint:disable=C3001
+
+        page = iamraw.Page(ltpage.pageid, iamraw.BoundingBox(*ltpage.bbox))
+
+        for item in chars:
+            page.append(item)
+        self.document.pages.append(page)  # pylint:disable=E1101

rawmaker/miner/colorspace.py

@@ -0,0 +1,75 @@
+# =============================================================================
+# C O P Y R I G H T
+# -----------------------------------------------------------------------------
+# Copyright (c) 2021-2023 by Helmut Konrad Schewe. All rights reserved.
+# This file is property of Helmut Konrad Schewe. Any unauthorized copy,
+# use or distribution is an offensive act against international law and may
+# be prosecuted under federal law. Its content is company confidential.
+# =============================================================================
+
+import contextlib
+
+import rawmaker.utils
+
+
+def parse(colorspace) -> str:  # pylint:disable=R0911
+    """\
+    >>> import pdfminer.psparser
+    >>> parse([pdfminer.psparser.PSLiteral('CS')]) # verify later
+    'CS'
+    """
+    if len(colorspace) == 1:
+        colorspace = colorspace[0]
+    if not colorspace:
+        # TODO: VERIFY THIS
+        return 'DeviceGray'
+    colorspace = name(colorspace)
+    if isinstance(colorspace, str):
+        # TODO: VERIFY R212!
+        if colorspace in {'DeviceRGB', 'RGB', 'R213'}:
+            # RGB is an abbreviation of DeviceRGB
+            return 'DeviceRGB'
+        if colorspace in {'DeviceGray', 'G'}:
+            # G is an abbreviation of DeviceGray
+            return 'DeviceGray'
+        if 'CS' in colorspace:
+            return colorspace
+        return colorspace
+    typ = colorspace[0].name
+    if typ == 'Indexed':
+        return indexed_space(*colorspace[1:])
+    if typ == 'ICCBased':
+        return iccbased(rawmaker.utils.resolve(colorspace[1]))
+    return 'DeviceRGB'
+
+
+def indexed_space(base, hival, lookup):  # pylint:disable=W0613
+    base = name(base)
+    if base[0] == 'ICCBased':
+        return iccbased(rawmaker.utils.resolve(base[1]))
+    if str(base) == 'DeviceRGB':
+        return 'DeviceRGB'
+    return None
+
+
+def iccbased(stream) -> str:
+    attributes = stream.attrs
+    # rawdata = stream.rawdata  # pylint:disable=W0612
+    with contextlib.suppress(KeyError):
+        colorspace = attributes['N']
+        if colorspace == 1:
+            colorspace = 'DeviceGray'
+        elif colorspace == 3:
+            colorspace = 'DeviceRGB'
+        elif colorspace == 4:
+            colorspace = 'DeviceCMYK'
+    with contextlib.suppress(KeyError):
+        colorspace = attributes['Alternate'].name
+    return colorspace
+
+
+def name(reference) -> str:
+    reference = rawmaker.utils.resolve(reference)
+    with contextlib.suppress(AttributeError):
+        return reference.name
+    return reference