Skip to content

Commit

Permalink
Implement source spans for table nodes
Browse files Browse the repository at this point in the history
  • Loading branch information
@robinst
robinst committed Sep 7, 2020
1 parent 149ed27 commit e86d5ab
Show file tree
Hide file tree
Showing 2 changed files with 81 additions and 37 deletions.
112 changes: 78 additions & 34 deletions ...ext-gfm-tables/src/main/java/org/commonmark/ext/gfm/tables/internal/TableBlockParser.java
View file
Original file line number Diff line number Diff line change
@@ -1,26 +1,34 @@
package org.commonmark.ext.gfm.tables.internal;

import org.commonmark.ext.gfm.tables.*;
import org.commonmark.ext.gfm.tables.TableBlock;
import org.commonmark.ext.gfm.tables.TableBody;
import org.commonmark.ext.gfm.tables.TableCell;
import org.commonmark.ext.gfm.tables.TableHead;
import org.commonmark.ext.gfm.tables.TableRow;
import org.commonmark.node.Block;
import org.commonmark.node.Node;
import org.commonmark.node.SourceSpan;
import org.commonmark.parser.InlineParser;
import org.commonmark.parser.block.*;
import org.commonmark.parser.block.AbstractBlockParser;
import org.commonmark.parser.block.AbstractBlockParserFactory;
import org.commonmark.parser.block.BlockContinue;
import org.commonmark.parser.block.BlockStart;
import org.commonmark.parser.block.MatchedBlockParser;
import org.commonmark.parser.block.ParserState;

import java.util.ArrayList;
import java.util.Collections;
import java.util.List;

public class TableBlockParser extends AbstractBlockParser {

private final TableBlock block = new TableBlock();
private final List<CharSequence> bodyLines = new ArrayList<>();
private final List<CharSequence> rowLines = new ArrayList<>();
private final List<TableCell.Alignment> columns;
private final List<String> headerCells;

private boolean nextIsSeparatorLine = true;

private TableBlockParser(List<TableCell.Alignment> columns, List<String> headerCells) {
private TableBlockParser(List<TableCell.Alignment> columns, CharSequence headerLine) {
this.columns = columns;
this.headerCells = headerCells;
this.rowLines.add(headerLine);
}

@Override
Expand All @@ -44,37 +52,45 @@ public BlockContinue tryContinue(ParserState state) {

@Override
public void addLine(CharSequence line) {
if (nextIsSeparatorLine) {
nextIsSeparatorLine = false;
} else {
bodyLines.add(line);
}
rowLines.add(line);
}

@Override
public void parseInlines(InlineParser inlineParser) {
int headerColumns = headerCells.size();
List<SourceSpan> sourceSpans = block.getSourceSpans();

SourceSpan headerSourceSpan = !sourceSpans.isEmpty() ? sourceSpans.get(0) : null;
Node head = new TableHead();
if (headerSourceSpan != null) {
head.getSourceSpans().add(headerSourceSpan);
}
block.appendChild(head);

TableRow headerRow = new TableRow();
headerRow.setSourceSpans(head.getSourceSpans());
head.appendChild(headerRow);

List<CellSource> headerCells = split(rowLines.get(0), headerSourceSpan);
int headerColumns = headerCells.size();
for (int i = 0; i < headerColumns; i++) {
String cell = headerCells.get(i);
CellSource cell = headerCells.get(i);
TableCell tableCell = parseCell(cell, i, inlineParser);
tableCell.setHeader(true);
headerRow.appendChild(tableCell);
}

Node body = null;
for (CharSequence rowLine : bodyLines) {
List<String> cells = split(rowLine);
TableBody body = null;
// Body starts at index 2. 0 is header, 1 is separator.
for (int rowIndex = 2; rowIndex < rowLines.size(); rowIndex++) {
CharSequence rowLine = rowLines.get(rowIndex);
SourceSpan sourceSpan = sourceSpans.get(rowIndex);
List<CellSource> cells = split(rowLine, sourceSpan);
TableRow row = new TableRow();
row.getSourceSpans().add(sourceSpan);

// Body can not have more columns than head
for (int i = 0; i < headerColumns; i++) {
String cell = i < cells.size() ? cells.get(i) : "";
CellSource cell = i < cells.size() ? cells.get(i) : new CellSource("", null);
TableCell tableCell = parseCell(cell, i, inlineParser);
row.appendChild(tableCell);
}
Expand All @@ -85,33 +101,35 @@ public void parseInlines(InlineParser inlineParser) {
block.appendChild(body);
}
body.appendChild(row);
body.getSourceSpans().add(sourceSpan);
}
}

private TableCell parseCell(String cell, int column, InlineParser inlineParser) {
private TableCell parseCell(CellSource cell, int column, InlineParser inlineParser) {
TableCell tableCell = new TableCell();

if (column < columns.size()) {
tableCell.setAlignment(columns.get(column));
}

inlineParser.parse(cell.trim(), tableCell);
if (cell.sourceSpan != null) {
tableCell.setSourceSpans(Collections.singletonList(cell.sourceSpan));
}

inlineParser.parse(cell.content, tableCell);

return tableCell;
}

private static List<String> split(CharSequence input) {
String line = input.toString().trim();
if (line.startsWith("|")) {
line = line.substring(1);
}
List<String> cells = new ArrayList<>();
private static List<CellSource> split(CharSequence row, SourceSpan rowSourceSpan) {
int cellStart = row.charAt(0) == '|' ? 1 : 0;

This comment has been minimized.

Sign in to view

Copy link
@andreyxebialabs

andreyxebialabs Oct 2, 2021

Hello Robin, sorry for bothering about a commit made a year ago. The behavior changed here, because .trim() was removed, and now, if table has any trailing spaces before \n, these trailing spaces are considered as an incomplete cell and table is not recognized as a table. I'm not sure if this change was on purpose (e.g. because of enforcing strict formatting requirements) or not.

This comment has been minimized.

Sign in to view

Copy link
@robinst

robinst Oct 5, 2021

Author Collaborator

Hey! Yeah no worries. I think this was fixed by #244

Can you have a look?
List<CellSource> cells = new ArrayList<>();
StringBuilder sb = new StringBuilder();
for (int i = 0; i < line.length(); i++) {
char c = line.charAt(i);
for (int i = cellStart; i < row.length(); i++) {
char c = row.charAt(i);
switch (c) {
case '\\':
if (i + 1 < line.length() && line.charAt(i + 1) == '|') {
if (i + 1 < row.length() && row.charAt(i + 1) == '|') {
// Pipe is special for table parsing. An escaped pipe doesn't result in a new cell, but is
// passed down to inline parsing as an unescaped pipe. Note that that applies even for the `\|`
// in an input like `\\|` - in other words, table parsing doesn't support escaping backslashes.
Expand All @@ -123,15 +141,19 @@ private static List<String> split(CharSequence input) {
}
break;
case '|':
cells.add(sb.toString());
String content = sb.toString();
cells.add(CellSource.of(content, rowSourceSpan, cellStart));
sb.setLength(0);
// + 1 to skip the pipe itself for the next cell's span
cellStart = i + 1;
break;
default:
sb.append(c);
}
}
if (sb.length() > 0) {
cells.add(sb.toString());
String content = sb.toString();
cells.add(CellSource.of(content, rowSourceSpan, cellStart));
}
return cells;
}
Expand Down Expand Up @@ -229,9 +251,9 @@ public BlockStart tryStart(ParserState state, MatchedBlockParser matchedBlockPar
CharSequence separatorLine = line.subSequence(state.getIndex(), line.length());
List<TableCell.Alignment> columns = parseSeparator(separatorLine);
if (columns != null && !columns.isEmpty()) {
List<String> headerCells = split(paragraph);
List<CellSource> headerCells = split(paragraph, null);
if (columns.size() >= headerCells.size()) {
return BlockStart.of(new TableBlockParser(columns, headerCells))
return BlockStart.of(new TableBlockParser(columns, paragraph))
.atIndex(state.getIndex())
.replaceActiveBlockParser();
}
Expand All @@ -241,4 +263,26 @@ public BlockStart tryStart(ParserState state, MatchedBlockParser matchedBlockPar
}
}

private static class CellSource {

private final String content;
private final SourceSpan sourceSpan;

public static CellSource of(String content, SourceSpan rowSourceSpan, int offset) {
if (!content.isEmpty()) {
SourceSpan sourceSpan = null;
if (rowSourceSpan != null) {
sourceSpan = SourceSpan.of(rowSourceSpan.getLineIndex(), rowSourceSpan.getColumnIndex() + offset, content.length());
}
return new CellSource(content, sourceSpan);
} else {
return new CellSource(content, null);
}
}

private CellSource(String content, SourceSpan sourceSpan) {
this.content = content;
this.sourceSpan = sourceSpan;
}
}
}
6 changes: 3 additions & 3 deletions commonmark-ext-gfm-tables/src/test/java/org/commonmark/ext/gfm/tables/TablesTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -666,7 +666,7 @@ public void sourceSpans() {
assertEquals(Arrays.asList(SourceSpan.of(0, 4, 3)), headRowCell2.getSourceSpans());

TableBody body = (TableBody) block.getLastChild();
assertEquals(Arrays.asList(SourceSpan.of(2, 0, 4), SourceSpan.of(3, 1, 7), SourceSpan.of(4, 0, 3)), body.getSourceSpans());
assertEquals(Arrays.asList(SourceSpan.of(2, 0, 4), SourceSpan.of(3, 0, 8), SourceSpan.of(4, 0, 3)), body.getSourceSpans());

TableRow bodyRow1 = (TableRow) body.getFirstChild();
assertEquals(Arrays.asList(SourceSpan.of(2, 0, 4)), bodyRow1.getSourceSpans());
Expand All @@ -676,10 +676,10 @@ public void sourceSpans() {
assertEquals(Arrays.asList(SourceSpan.of(2, 3, 1)), bodyRow1Cell2.getSourceSpans());

TableRow bodyRow2 = (TableRow) body.getFirstChild().getNext();
assertEquals(Arrays.asList(SourceSpan.of(3, 1, 7)), bodyRow2.getSourceSpans());
assertEquals(Arrays.asList(SourceSpan.of(3, 0, 8)), bodyRow2.getSourceSpans());
TableCell bodyRow2Cell1 = (TableCell) bodyRow2.getFirstChild();
TableCell bodyRow2Cell2 = (TableCell) bodyRow2.getLastChild();
assertEquals(Arrays.asList(SourceSpan.of(3, 1, 1)), bodyRow2Cell1.getSourceSpans());
assertEquals(Arrays.asList(SourceSpan.of(3, 0, 2)), bodyRow2Cell1.getSourceSpans());
assertEquals(Arrays.asList(SourceSpan.of(3, 3, 4)), bodyRow2Cell2.getSourceSpans());

TableRow bodyRow3 = (TableRow) body.getLastChild();
Expand Down

0 comments on commit e86d5ab

Please sign in to comment.