Render embedded plain footnotes

src/bibleref.nim

@@ -98,6 +98,21 @@ proc fetchPassages(reference, translation: string, cfg: CombinedConfig): seq[str
       "unsupported translation '" & translation &
         "'; supported translations: " & supportedTranslationsList())
 
+proc fetchPlainPassages(
+    reference,
+    translation: string,
+    keepVerseNumbers: bool,
+    cfg: CombinedConfig): seq[string] =
+
+  case translation
+  of "akjv", "kjv":
+    result = kjv.fetchPlainPassages(reference, keepVerseNumbers)
+  of "mev":
+    result = mev.fetchPlainPassages(reference, keepVerseNumbers)
+  else:
+    for passage in fetchPassages(reference, translation, cfg):
+      result.add(formatPlain(passage, translation, keepVerseNumbers))
+
 when isMainModule:
   const USAGE = """Usage:
   bibleref <reference> [options]
@@ -167,11 +182,17 @@ Options:
 
     var formattedPassages: seq[string] = @[]
     for query in queries:
+      case $args["--output-format"]
+      of "plain":
+        formattedPassages.add(fetchPlainPassages(
+          query.referenceText,
+          query.translation,
+          keepVerseNumbers = true,
+          cfg))
+      else:
         for passage in fetchPassages(query.referenceText, query.translation, cfg):
           formattedPassages.add(
             case $args["--output-format"]:
-          of "plain":
-            formatPlain(passage, query.translation)
             of "reading":
               formatPlain(passage, query.translation, keepVerseNumbers = false)
             of "text":

src/embedded_bible.nim

@@ -1,4 +1,4 @@
-import std/[strutils, tables]
+import std/[algorithm, json, strutils, tables, wordwrap]
 
 import ./reference_parser
 
@@ -16,6 +16,12 @@ type
     breakKind: SegmentBreak
     indent: int
     text: string
+    annotations: seq[BibleAnnotation]
+
+  BibleAnnotation = object
+    kind: string
+    offset: int
+    targetText: string
 
   BibleIndex = object
     segments: seq[BibleSegment]
@@ -49,6 +55,31 @@ proc parseSegmentBreak(s: string): SegmentBreak =
   else:
     raise newException(ValueError, "invalid embedded Bible segment break: " & s)
 
+proc parseAnnotations(metadata: string): seq[BibleAnnotation] =
+  if metadata.strip.len == 0:
+    return
+
+  let node = parseJson(metadata)
+  if node.kind != JObject or not node.hasKey("annotations"):
+    return
+
+  for item in node["annotations"].getElems:
+    let targetText =
+      if item.hasKey("targetText"): item["targetText"].getStr
+      else: ""
+
+    if targetText.strip.len == 0:
+      continue
+
+    result.add(BibleAnnotation(
+      kind:
+        if item.hasKey("kind"): item["kind"].getStr
+        else: "",
+      offset:
+        if item.hasKey("offset"): item["offset"].getInt
+        else: 0,
+      targetText: targetText))
+
 proc addSegment(index: var BibleIndex, segment: BibleSegment) =
   index.segments.add(segment)
   index.verses[verseKey(segment.code, segment.chapter, segment.verse)] = true
@@ -92,7 +123,10 @@ proc loadBibleIndex(rows, translationName: string): BibleIndex =
         segment: parseInt(parts[3]),
         breakKind: parseSegmentBreak(parts[4]),
         indent: parseInt(parts[5]),
-        text: parts[6]))
+        text: parts[6],
+        annotations:
+          if parts.len == 8: parseAnnotations(parts[7])
+          else: @[]))
 
 proc requireLastChapter(index: BibleIndex, code: string): int =
   if not index.lastChapterByBook.hasKey(code):
@@ -245,3 +279,174 @@ proc fetchPassages*(rows, reference, translationName: string): seq[string] =
   let index = loadBibleIndex(rows, translationName)
   for parsedReference in parseReferences(reference):
     result.add(fetchReference(index, parsedReference))
+
+proc footnoteLabel(index: int): string =
+  var n = index
+  while true:
+    result.insert($char(ord('a') + (n mod 26)), 0)
+    n = (n div 26) - 1
+    if n < 0:
+      break
+
+proc addSegmentWithFootnotes(
+    line: var string,
+    segment: BibleSegment,
+    keepVerseNumbers: bool,
+    notes: var seq[string]) =
+
+  var text = segment.text
+  var annotations = segment.annotations
+  annotations.sort(proc(a, b: BibleAnnotation): int = cmp(a.offset, b.offset))
+
+  var inserted = 0
+  for annotation in annotations:
+    let label = footnoteLabel(notes.len)
+    notes.add(annotation.targetText)
+
+    let insertAt = max(0, min(text.len, annotation.offset + inserted))
+    let marker = "[" & label & "]"
+    text.insert(marker, insertAt)
+    inserted += marker.len
+
+  let marker =
+    if keepVerseNumbers: $segment.verse & " "
+    else: ""
+
+  let content = marker & text
+  if line.len == 0:
+    line.add(repeat("  ", segment.indent) & content)
+  elif segment.breakKind == sbSame:
+    line.appendSameLine(content)
+  else:
+    line.add("\n" & repeat("  ", segment.indent) & content)
+
+proc addFootnotes(lines: var seq[string], notes: seq[string]) =
+  if notes.len == 0:
+    return
+
+  lines.add("")
+  for idx, note in notes:
+    lines.add(footnoteLabel(idx) & ". " & note)
+
+proc selectedChapters(index: BibleIndex, reference: PassageReference):
+    seq[tuple[chapter: int, verses: Table[string, bool]]] =
+
+  let code = reference.book.code
+
+  if reference.ranges.len == 0:
+    for chapter in 1 .. index.requireLastChapter(code):
+      var verses: Table[string, bool]
+      for verse in 1 .. index.requireLastVerse(code, chapter):
+        discard index.requireVerse(code, chapter, verse)
+        verses[verseKey(code, chapter, verse)] = true
+      result.add((chapter, verses))
+  else:
+    var chapterPositions: Table[int, int]
+    for range in reference.ranges:
+      let selected = index.selectedVerses(reference, range)
+      for key in selected.keys:
+        let parts = key.split('\t')
+        let chapter = parseInt(parts[1])
+        if not chapterPositions.hasKey(chapter):
+          chapterPositions[chapter] = result.len
+          var verses: Table[string, bool]
+          result.add((chapter, verses))
+        result[chapterPositions[chapter]].verses[key] = true
+
+proc plainChapter(
+    index: BibleIndex,
+    code: string,
+    selected: Table[string, bool],
+    keepVerseNumbers: bool): string =
+
+  var lines: seq[string]
+  var currentLine = ""
+  var notes: seq[string]
+  var emittedVerses: Table[string, bool]
+  var firstSelected = true
+
+  proc flushLine() =
+    if currentLine.len > 0:
+      lines.add(currentLine)
+      currentLine = ""
+
+  for segment in index.segments:
+    if segment.code != code:
+      continue
+
+    let key = verseKey(segment.code, segment.chapter, segment.verse)
+    if not selected.hasKey(key):
+      continue
+
+    var adjusted = segment
+    if firstSelected:
+      adjusted = normalizeFirstSelectedSegment(adjusted)
+
+    let startsNewParagraph =
+      adjusted.breakKind == sbParagraph and
+      not firstSelected and
+      (not emittedVerses.hasKey(key))
+
+    if startsNewParagraph:
+      flushLine()
+      lines.add("")
+    elif adjusted.breakKind == sbLine:
+      flushLine()
+
+    let marker =
+      if emittedVerses.hasKey(key): false
+      else:
+        emittedVerses[key] = true
+        true
+
+    var renderSegment = adjusted
+    if not marker:
+      renderSegment.verse = 0
+
+    if marker:
+      currentLine.addSegmentWithFootnotes(renderSegment, keepVerseNumbers, notes)
+    else:
+      var noVerseSegment = renderSegment
+      noVerseSegment.text = renderSegment.text
+      noVerseSegment.breakKind = sbSame
+      currentLine.addSegmentWithFootnotes(noVerseSegment, false, notes)
+
+    firstSelected = false
+
+  flushLine()
+  lines.addFootnotes(notes)
+
+  var wrapped: seq[string]
+  for line in lines:
+    if line.len == 0:
+      wrapped.add("")
+      continue
+
+    let prepared =
+      if line.len > 90: line.strip
+      else: line & "  "
+    wrapped.add(wrapWords(prepared, maxLineWidth = 74, newLine = "\p"))
+  wrapped.join("\p")
+
+proc fetchPlainPassages*(
+    rows,
+    reference,
+    translationName: string,
+    keepVerseNumbers = true): seq[string] =
+
+  let index = loadBibleIndex(rows, translationName)
+  for parsedReference in parseReferences(reference):
+    var parts: seq[string]
+    for chapter in index.selectedChapters(parsedReference):
+      parts.add(index.plainChapter(
+        parsedReference.book.code,
+        chapter.verses,
+        keepVerseNumbers))
+
+    let body = parts.join("\p\p")
+    let referenceSeparator =
+      if body.contains("\pa. "): "\p\p"
+      else: "\p"
+
+    result.add(body & referenceSeparator & "– " & $parsedReference & " (" &
+      translationName.toUpperAscii & ")")

src/kjv.nim

@@ -5,3 +5,6 @@ const kjvRows = embeddedTranslationData("kjv")
 
 proc fetchPassages*(reference: string): seq[string] =
   embedded_bible.fetchPassages(kjvRows, reference, "KJV")
+
+proc fetchPlainPassages*(reference: string, keepVerseNumbers = true): seq[string] =
+  embedded_bible.fetchPlainPassages(kjvRows, reference, "KJV", keepVerseNumbers)

src/mev.nim

@@ -11,3 +11,10 @@ proc fetchPassages*(reference: string): seq[string] =
   else:
     raise newException(ValueError,
       "MEV data is not embedded; generate data/private/mev.tsv and rebuild")
+
+proc fetchPlainPassages*(reference: string, keepVerseNumbers = true): seq[string] =
+  when hasEmbeddedTranslationData("mev"):
+    embedded_bible.fetchPlainPassages(mevRows, reference, "MEV", keepVerseNumbers)
+  else:
+    raise newException(ValueError,
+      "MEV data is not embedded; generate data/private/mev.tsv and rebuild")

tests/test_offline_kjv.nim

@@ -170,3 +170,37 @@ HEB	4	7	2	line	1	do not harden your hearts."
       "  [31] Then Jesus said. [32] You shall know the truth.\n" &
       "\n" &
       "  [33] They answered Him."
+
+  test "plain output renders annotation footnotes":
+    const rows =
+      "JHN\t8\t1\t0\tparagraph\t0\tBut Jesus went to the Mount of Olives.\t" &
+      """{"annotations":[{"kind":"crossReference","marker":"a","offset":40,"targetText":"Mt 21:1"}]}""" & "\n" &
+      "JHN\t8\t2\t0\tsame\t0\tEarly in the morning He returned to the temple.\t" &
+      """{"annotations":[{"kind":"crossReference","marker":"a","offset":49,"targetText":"Mt 26:55; Lk 4:20"}]}""" & "\n"
+    let passages = embedded_bible.fetchPlainPassages(rows, "John 8:1-2", "TEST")
+
+    check passages[0] == "1 But Jesus went to the Mount of Olives.[a] 2 Early in the morning He\n" &
+      "returned to the temple.[b]\n" &
+      "\n" &
+      "a. Mt 21:1\n" &
+      "b. Mt 26:55; Lk 4:20\n" &
+      "\n" &
+      "– John 8:1-2 (TEST)"
+
+  test "plain output resets footnote labels per chapter":
+    const rows =
+      "JHN\t8\t59\t0\tparagraph\t0\tThey took up stones.\t" &
+      """{"annotations":[{"kind":"crossReference","marker":"a","offset":21,"targetText":"Lev 24:16"}]}""" & "\n" &
+      "JHN\t9\t1\t0\tparagraph\t0\tAs Jesus passed by.\t" &
+      """{"annotations":[{"kind":"crossReference","marker":"a","offset":20,"targetText":"Jn 8:12"}]}""" & "\n"
+    let passages = embedded_bible.fetchPlainPassages(rows, "John 8:59-9:1", "TEST")
+
+    check passages[0] == "59 They took up stones.[a]\n" &
+      "\n" &
+      "a. Lev 24:16\n" &
+      "\n" &
+      "1 As Jesus passed by.[a]\n" &
+      "\n" &
+      "a. Jn 8:12\n" &
+      "\n" &
+      "– John 8:59-9:1 (TEST)"

tools/generate_mev_data.nim

@@ -273,15 +273,24 @@ proc annotationKind(marker: string): string =
   if marker.strip.allCharsInSet({'0'..'9'}): "note"
   else: "crossReference"
 
+proc cleanTargetText(targetText: string): string =
+  var parts = targetText.splitWhitespace
+  if parts.len >= 3 and parts[0].allCharsInSet({'0'..'9'}):
+    parts.delete(0)
+    parts.delete(0)
+    return parts.join(" ")
+
+  targetText
+
 proc annotationNode(marker, href, targetText: string, offset: int): JsonNode =
+  let kind = annotationKind(marker)
+  let cleanedTargetText = cleanTargetText(targetText)
   result = %*{
-    "kind": annotationKind(marker),
-    "marker": marker,
-    "href": href,
+    "kind": kind,
     "offset": offset
   }
-  if targetText.len > 0:
-    result["targetText"] = newJString(targetText)
+  if cleanedTargetText.len > 0:
+    result["targetText"] = newJString(cleanedTargetText)
 
 proc segmentMetadata(state: ParseState): JsonNode =
   result = newJObject()