Preserve layout in embedded Bible data

tools/generate_kjv_data.nim

@@ -84,6 +84,22 @@ proc parseVerseLine(line: string): tuple[verse: int, text: string] =
   result.verse = parseInt(rest[0 ..< numberEnd])
   result.text = stripUsfmMarkup(rest[numberEnd + 1 .. ^1])
 
+proc usfmBlock(line: string): tuple[matched: bool, segmentBreak: string, indent: int] =
+  if line.startsWith("\\p") or line.startsWith("\\m"):
+    return (true, "paragraph", 0)
+
+  if line.startsWith("\\q"):
+    var idx = 2
+    var digits = ""
+    while idx < line.len and line[idx].isDigit:
+      digits.add(line[idx])
+      inc idx
+
+    let indent =
+      if digits.len > 0: parseInt(digits)
+      else: 1
+    return (true, "line", indent)
+
 proc findCanonFiles(inputDir: string): Table[string, string] =
   for path in walkFiles(inputDir / "*eng-kjv.usfm"):
     let name = path.extractFilename
@@ -104,35 +120,58 @@ proc generate(inputDir, outputPath: string) =
 
     var chapter = 0
     var verse = 0
-    var verseText = ""
+    var segment = 0
+    var segmentText = ""
+    var segmentBreak = "line"
+    var indent = 0
 
-    proc flushVerse() =
+    proc flushSegment() =
       if chapter > 0 and verse > 0:
-        let text = normalizeWhitespace(verseText).replace("\t", " ")
+        let text = normalizeWhitespace(segmentText).replace("\t", " ")
         if text.len > 0:
-          rows.add([code, $chapter, $verse, text].join("\t"))
-      verse = 0
-      verseText = ""
+          rows.add([
+            code,
+            $chapter,
+            $verse,
+            $segment,
+            segmentBreak,
+            $indent,
+            text
+          ].join("\t"))
+          inc segment
+          segmentBreak = "same"
+      segmentText = ""
+
+    proc startBlock(nextBreak: string, nextIndent: int) =
+      flushSegment()
+      segmentBreak = nextBreak
+      indent = nextIndent
 
     for rawLine in canonFiles[code].lines:
       let line = rawLine.strip
 
       if line.startsWith("\\c "):
-        flushVerse()
+        flushSegment()
         chapter = parseInt(line[3..^1].strip)
+        verse = 0
       elif line.startsWith("\\v "):
-        flushVerse()
+        flushSegment()
         let parsed = parseVerseLine(line)
         verse = parsed.verse
-        verseText = parsed.text
+        segment = 0
+        segmentText = parsed.text
       elif verse > 0:
+        let block = usfmBlock(line)
+        if block.matched:
+          startBlock(block.segmentBreak, block.indent)
+
         let continued = stripUsfmMarkup(line)
         if continued.len > 0:
-          if verseText.len > 0:
-            verseText.add(' ')
-          verseText.add(continued)
+          if segmentText.len > 0:
+            segmentText.add(' ')
+          segmentText.add(continued)
 
-    flushVerse()
+    flushSegment()
 
   createDir(outputPath.parentDir)
   writeFile(outputPath, rows.join("\n") & "\n")