Simplify rendering calculations.

Author: jzohrab

lute/book/stats.py

@@ -59,10 +59,7 @@ def flatten_list(nested_list):
                 result.append(item)
         return result
 
-    text_items = []
-    for s in flatten_list(paras):
-        text_items.extend(s.textitems)
-    text_items = [ti for ti in text_items if ti.is_word]
+    text_items = [ti for ti in flatten_list(paras) if ti.is_word]
 
     statterms = {0: [], 1: [], 2: [], 3: [], 4: [], 5: [], 98: [], 99: []}
 

lute/cli/language_term_export.py

@@ -63,7 +63,7 @@ def _process_book(b, terms):
             ti.term
             for para in paragraphs
             for sentence in para
-            for ti in sentence.textitems
+            for ti in sentence
             if ti.is_word and ti.term is not None
         ]
         for t in displayed_terms:

lute/read/render/calculate_textitems.py

@@ -0,0 +1,239 @@
+"""
+Given text and Terms, determine what to render in the browser.
+
+For example, given the following TextTokens A-I:
+
+  A B C D E F G H I
+
+And the following terms:
+
+  "A" through "I" (single-word terms)
+  "B C"       (term J)
+  "E F G H I" (K)
+  "F G"       (L)
+  "C D E"     (M)
+
+The following TextItems would be displayed on the reading screen,
+with some of the Terms overlapping:
+
+  [A][B C][-D E][-F G H I]
+"""
+
+import re
+from lute.models.term import Term
+from lute.read.render.text_item import TextItem
+
+zws = "\u200B"  # zero-width space
+
+
+def get_string_indexes(strings, content):
+    """
+    Returns list of arrays: [[string, index], ...]
+
+    e.g., _get_string_indexes(["is a", "cat"], "here is a cat")
+    returns [("is a", 1), ("cat", 3)].
+
+    strings and content must be lowercased!
+    """
+    searchcontent = zws + content + zws
+    zwsindexes = [index for index, letter in enumerate(searchcontent) if letter == zws]
+
+    ret = []
+
+    for s in strings:
+        # "(?=())" is required because sometimes the search pattern can
+        # overlap -- e.g. _b_b_ has _b_ *twice*.
+        # https://stackoverflow.com/questions/5616822/
+        #   how-to-use-regex-to-find-all-overlapping-matches
+        pattern = rf"(?=({re.escape(zws + s + zws)}))"
+        add_matches = [
+            (s, zwsindexes.index(m.start()))
+            for m in re.finditer(pattern, searchcontent)
+        ]
+        ret.extend(add_matches)
+
+    return ret
+
+
+def _make_textitem(index, text, text_lc, sentence_number, term):
+    "Make a TextItem."
+    r = TextItem()
+    r.text = text
+    r.sentence_number = sentence_number
+    r.text_lc = text_lc
+    r.token_count = text.count(zws) + 1
+    r.display_count = r.token_count
+    r.index = index
+    r.is_word = term is not None
+    r.term = term
+    return r
+
+
+def _create_missing_status_0_terms(tokens, terms, language):
+    "Make new terms as needed for all tokens, using case of last instance."
+
+    original_word_tokens = {t.token for t in tokens if t.is_word}
+    parser = language.parser
+    lc_word_tokens = {parser.get_lowercase(t): t for t in original_word_tokens}
+    term_text_lcs = {t.text_lc for t in terms}
+
+    missing_word_tokens = [
+        original for lc, original in lc_word_tokens.items() if lc not in term_text_lcs
+    ]
+
+    # Note: create the terms _without parsing_ because some parsers
+    # break up characters when the words are given out of context.
+    missing_word_tokens = list(set(missing_word_tokens))
+    new_terms = [Term.create_term_no_parsing(language, t) for t in missing_word_tokens]
+    for t in new_terms:
+        t.status = 0
+
+    return new_terms
+
+
+def get_textitems(tokens, terms, language):
+    """
+    Return TextItems that will **actually be rendered**.
+
+    Method to determine what should be rendered:
+
+    - Create TextItems for all of the tokens, finding their
+      starting index in the tokens.
+
+    - "Write" the TextItems to an array in correctly sorted
+      order, so that the correct TextItems take precendence
+      in the final rendering.
+
+    - Calculate any term overlaps.
+
+    - Return the final list of TextItems that will actually
+      be rendered.
+
+    ---
+
+    Applying the above algorithm to the example given in the class
+    header:
+
+    We have the following TextTokens A-I:
+
+      A B C D E F G H I
+
+    And given the following terms:
+      "A" through "I" (single-word terms)
+      "B C"       (term J)
+      "E F G H I" (K)
+      "F G"       (L)
+      "C D E"     (M)
+
+    Creating TextItems for all of the terms, finding their starting
+    indices in the tokens:
+
+      TextToken    index   length
+      ----         -----   ------
+      [A]          0       1
+      [B]          1       1
+      ...
+      [I]          8       1
+      [B C]        1       2
+      [E F G H I]  4       5
+      [F G]        5       2
+      [C D E]      2       3
+
+    Sorting by index, then decreasing token count:
+
+      TextToken    index   length   ID (for later reference)
+      ----         -----   ------   ------------------------
+      [A]          0       1        t1
+      [B C]        1       2        t2
+      [B]          1       1        t3
+      [C D E]      2       3        t4
+      [C]          2       1        t5
+      [D]          3       1        t6
+      [E F G H I]  4       5        t7
+      [E]          4       1        t8
+      [F G]        5       2        t9
+      [F]          5       1        t10
+      [G]          6       1        t11
+      [H]          7       1        t12
+      [I]          8       1        t13
+
+    Starting at the bottom of the above list and
+    working upwards:
+
+      - ID of [I] is written to index 8: [] [] [] [] [] [] [] [] [t13]
+      - ID of [H] to index 7: [] [] [] [] [] [] [] [t12] [t13]
+      - ...
+      - [F G] to index 5 *and* 6: [] [] [] [] [] [t9] [t9] [t12] [t13]
+      - [E] to index 4:  [] [] [] [] [t8] [t9] [t9] [t12] [t13]
+      - [E F G H I] to indexes 4-8:   [] [] [] [] [t7] [t7] [t7] [t7] [t7]
+      - ... etc
+
+    Using the TextItem IDs, the resulting array would be:
+
+      output array: [t1] [t2] [t2] [t4] [t4] [t7] [t7] [t7] [t7]
+                    [A]  [B     C] [-D    E] [-F    G    H    I]
+
+    The only TextItems that will be shown are therefore:
+      t1, t2, t3, t7
+
+    To calculate what text is actually displayed, the count
+    of each ID is used.  e.g.:
+      - ID t7 appears 4 times in the output array.  The last 4 tokens of
+        [E F G H I] are [F G H I], which will be used as t7's display text.
+      - ID t2 appears 2 times.  The last 2 tokens of [B C] are [B C],
+        so that will be the display text. etc.
+    """
+    # pylint: disable=too-many-locals
+
+    new_unknown_terms = _create_missing_status_0_terms(tokens, terms, language)
+
+    all_terms = terms + new_unknown_terms
+
+    text_to_term = {dt.text_lc: dt for dt in all_terms}
+
+    tokens_lc = [language.parser.get_lowercase(t.token) for t in tokens]
+
+    textitems = []
+
+    def _add_textitem(index, text_lc):
+        "Add a TextItem for position index in tokens."
+        count = text_lc.count(zws) + 1
+        text_orig = zws.join([t.token for t in tokens[index : index + count]])
+        text_lc = zws.join(tokens_lc[index : index + count])
+        sentence_number = tokens[index].sentence_number
+        term = text_to_term.get(text_lc, None)
+        ti = _make_textitem(index, text_orig, text_lc, sentence_number, term)
+        textitems.append(ti)
+
+    # Single-word terms.
+    for index, _ in enumerate(tokens):
+        _add_textitem(index, tokens_lc[index])
+
+    # Multiword terms.
+    multiword_terms = [t.text_lc for t in all_terms if t.token_count > 1]
+    for e in get_string_indexes(multiword_terms, zws.join(tokens_lc)):
+        _add_textitem(e[1], e[0])
+
+    # Sorting by index, then decreasing token count.
+    textitems = sorted(textitems, key=lambda x: (x.index, -x.token_count))
+
+    # "Write out" TextItems to the output array.
+    output_textitem_ids = [None] * len(tokens)
+    for ti in reversed(textitems):
+        for c in range(ti.index, ti.index + ti.token_count):
+            output_textitem_ids[c] = id(ti)
+
+    # Calc display_counts; e.g. if a textitem's id shows up 3 times
+    # in the output_textitem_ids, it should display 3 tokens.
+    for ti in textitems:
+        ti.display_count = output_textitem_ids.count(id(ti))
+
+    textitems = [ti for ti in textitems if ti.display_count > 0]
+
+    current_paragraph = 0
+    for ti in textitems:
+        ti.paragraph_number = current_paragraph
+        if ti.text == "¶":
+            current_paragraph += 1
+
+    return textitems