Skip to content
This repository has been archived by the owner on May 7, 2024. It is now read-only.

Commit

Permalink
Determining the headings.
Browse files Browse the repository at this point in the history
  • Loading branch information
@walter-weinmann
walter-weinmann committed Jun 22, 2022
1 parent e281cf2 commit a1dc9cd
Showing 1 changed file with 24 additions and 56 deletions.
80 changes: 24 additions & 56 deletions src/dcr/nlp/cls_line_type_heading.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -88,17 +88,15 @@ def __init__(self) -> None:
# 2: is_first_token:
# True: apply rule to first token (split)
# False: apply rule to beginning of line
# 3: regexp_compiled_match:
# compiled regular expression for matching
# 4: regexp_compiled_display:
# compiled regular expression for showing the match
# 5: function_is_asc:
# 3: regexp_compiled:
# compiled regular expression
# 4: function_is_asc:
# compares predecessor and successor
# 6: start_values:
# list of strings
# -----------------------------------------------------------------------------
self._heading_rules_collection: list[
tuple[str, bool, re.Pattern[str], re.Pattern[str], collections.abc.Callable[[str, str], bool], list[str]]
tuple[str, bool, re.Pattern[str], collections.abc.Callable[[str, str], bool], list[str]]
] = []

for (rule_name, is_first_token, regexp_str, function_is_asc, start_values) in self._heading_rules:
Expand All @@ -107,7 +105,6 @@ def __init__(self) -> None:
rule_name.ljust(self._RULE_NAME_SIZE),
is_first_token,
re.compile(regexp_str),
re.compile(regexp_str + " [^.]+"),
function_is_asc,
start_values,
)
Expand All @@ -120,27 +117,24 @@ def __init__(self) -> None:
# 2: is_first_token:
# True: apply rule to first token (split)
# False: apply rule to beginning of line
# 3: regexp_compiled_match:
# compiled regular expression for matching
# 4: regexp_compiled_display:
# compiled regular expression for showing the match
# 5: function_is_asc:
# 3: regexp_compiled:
# compiled regular expression
# 4: function_is_asc:
# compares predecessor and successor
# 6: start_values:
# 5: start_values:
# list of strings
# 7: level:
# 6: level:
# hierarchical level of the current heading
# 8: lower_left_x:
# 7: lower_left_x:
# lower left x-coordinate of the beginning of the possible heading
# 9: predecessor:
# 8: predecessor:
# predecessor value
# -----------------------------------------------------------------------------
self._heading_rules_hierarchy: list[
tuple[
str,
bool,
re.Pattern[str],
re.Pattern[str],
collections.abc.Callable[[str, str], bool],
list[str],
int,
Expand Down Expand Up @@ -209,44 +203,22 @@ def _convert_roman_2_int(cls, roman: str) -> int:
# Create a table of content entry.
# -----------------------------------------------------------------------------
# pylint: disable=missing-param-doc
def _create_toc_entry(self, level: int, pattern_display: re.Match[str] | None, pattern_matched: re.Match[str]) -> str:
def _create_toc_entry(self, level: int, text: str) -> None:

"""Create a table of content entry.
Args:
level (int): Heading level.
pattern_display (re.Match[str] | None): Hits for display.
pattern_matched (re.Match[str] | None): Hits from search.
Returns:
str: The heading text.
text: Heading text.
"""
if not cfg.glob.setup.is_heading_create_toc:
return ""

if pattern_display is None:
heading = pattern_matched.group(0)
if self._idx_line_line < self._max_line_line:
heading = (
heading
+ ("" if heading[-1] == " " else " ")
+ cfg.glob.text_parser.parse_result_line_lines[self._idx_line_line + 1][
nlp.cls_nlp_core.NLPCore.JSON_NAME_TEXT
]
)
else:
heading = pattern_display.group(0)

self._toc.append(
{
nlp.cls_nlp_core.NLPCore.JSON_NAME_HEADING_LEVEL: level,
nlp.cls_nlp_core.NLPCore.JSON_NAME_HEADING_TEXT: heading,
nlp.cls_nlp_core.NLPCore.JSON_NAME_HEADING_TEXT: text,
nlp.cls_nlp_core.NLPCore.JSON_NAME_PAGE_NO: self._page_no,
}
)

return heading

# -----------------------------------------------------------------------------
# Initialise the heading rules.
# -----------------------------------------------------------------------------
Expand Down Expand Up @@ -599,8 +571,7 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901
(
rule_name,
is_first_token,
regexp_compiled_match,
regexp_compiled_display,
regexp_compiled,
function_is_asc,
start_values,
level,
Expand All @@ -610,7 +581,7 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901

target_value = first_token if is_first_token else text

if pattern_matched := regexp_compiled_match.match(target_value):
if regexp_compiled.match(target_value):
lower_left_x_curr_float = float(lower_left_x_curr)
lower_left_x_float = float(lower_left_x)
if (
Expand All @@ -623,8 +594,7 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901
self._heading_rules_hierarchy[ph_idx] = (
rule_name,
is_first_token,
regexp_compiled_match,
regexp_compiled_display,
regexp_compiled,
function_is_asc,
start_values,
level,
Expand All @@ -634,11 +604,11 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901

self._level_prev = level

heading = self._create_toc_entry(level, regexp_compiled_display.match(text), pattern_matched)
self._create_toc_entry(level, text)

utils.progress_msg_line_type_heading(
f"LineTypeHeading: Match ={rule_name} "
+ f"- level={level} - heading={heading}"
+ f"- level={level} - heading={text}"
)

# Delete levels that are no longer needed
Expand All @@ -653,13 +623,12 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901
for (
rule_name,
is_first_token,
regexp_compiled_match,
regexp_compiled_display,
regexp_compiled,
function_is_asc,
start_values,
) in self._heading_rules_collection:
target_value = first_token if is_first_token else text
if pattern_matched := regexp_compiled_match.match(target_value):
if regexp_compiled.match(target_value):
if is_first_token and start_values:
if first_token not in start_values:
continue
Expand All @@ -671,8 +640,7 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901
(
rule_name,
is_first_token,
regexp_compiled_match,
regexp_compiled_display,
regexp_compiled,
function_is_asc,
start_values,
level,
Expand All @@ -683,11 +651,11 @@ def _process_line(self, line_line: dict[str, str]) -> int: # noqa: C901

self._level_prev = level

heading = self._create_toc_entry(level, regexp_compiled_display.match(text), pattern_matched)
self._create_toc_entry(level, text)

utils.progress_msg_line_type_heading(
f"LineTypeHeading: Match new level ={rule_name} "
+ f"- level={level} - heading={heading}"
+ f"- level={level} - heading={text}"
)

return level
Expand Down

0 comments on commit a1dc9cd

Please sign in to comment.