diff options
| -rw-r--r-- | Hutch.xcodeproj/project.pbxproj | 16 | ||||
| -rw-r--r-- | Hutch/Views/Repositories/ReadmeView.swift | 699 | ||||
| -rw-r--r-- | HutchTests/ReadmeViewTests.swift | 211 |
3 files changed, 849 insertions, 77 deletions
diff --git a/Hutch.xcodeproj/project.pbxproj b/Hutch.xcodeproj/project.pbxproj index 347d6fc..a52d824 100644 --- a/Hutch.xcodeproj/project.pbxproj +++ b/Hutch.xcodeproj/project.pbxproj @@ -459,7 +459,7 @@ ASSETCATALOG_COMPILER_GLOBAL_ACCENT_COLOR_NAME = AccentColor; CODE_SIGN_ENTITLEMENTS = Hutch/Hutch.entitlements; CODE_SIGN_STYLE = Automatic; - CURRENT_PROJECT_VERSION = 11; + CURRENT_PROJECT_VERSION = 12; DEVELOPMENT_TEAM = ZCNAX3VL9D; ENABLE_PREVIEWS = YES; GENERATE_INFOPLIST_FILE = YES; @@ -476,7 +476,7 @@ "$(inherited)", "@executable_path/Frameworks", ); - MARKETING_VERSION = 2.4.1; + MARKETING_VERSION = 2.4.2; PRODUCT_BUNDLE_IDENTIFIER = net.cleberg.Hutch; PRODUCT_NAME = "$(TARGET_NAME)"; STRING_CATALOG_GENERATE_SYMBOLS = YES; @@ -496,7 +496,7 @@ ASSETCATALOG_COMPILER_GLOBAL_ACCENT_COLOR_NAME = AccentColor; CODE_SIGN_ENTITLEMENTS = Hutch/Hutch.entitlements; CODE_SIGN_STYLE = Automatic; - CURRENT_PROJECT_VERSION = 11; + CURRENT_PROJECT_VERSION = 12; DEVELOPMENT_TEAM = ZCNAX3VL9D; ENABLE_PREVIEWS = YES; GENERATE_INFOPLIST_FILE = YES; @@ -513,7 +513,7 @@ "$(inherited)", "@executable_path/Frameworks", ); - MARKETING_VERSION = 2.4.1; + MARKETING_VERSION = 2.4.2; PRODUCT_BUNDLE_IDENTIFIER = net.cleberg.Hutch; PRODUCT_NAME = "$(TARGET_NAME)"; STRING_CATALOG_GENERATE_SYMBOLS = YES; @@ -576,7 +576,7 @@ APPLICATION_EXTENSION_API_ONLY = YES; CODE_SIGN_ENTITLEMENTS = HutchWidgetExtension/HutchWidgetExtension.entitlements; CODE_SIGN_STYLE = Automatic; - CURRENT_PROJECT_VERSION = 11; + CURRENT_PROJECT_VERSION = 12; DEVELOPMENT_TEAM = ZCNAX3VL9D; GENERATE_INFOPLIST_FILE = NO; INFOPLIST_FILE = HutchWidgetExtension/Info.plist; @@ -586,7 +586,7 @@ "@executable_path/Frameworks", "@executable_path/../../Frameworks", ); - MARKETING_VERSION = 2.4.1; + MARKETING_VERSION = 2.4.2; PRODUCT_BUNDLE_IDENTIFIER = net.cleberg.Hutch.HutchWidgetExtension; PRODUCT_NAME = "$(TARGET_NAME)"; SKIP_INSTALL = YES; @@ -605,7 +605,7 @@ APPLICATION_EXTENSION_API_ONLY = YES; CODE_SIGN_ENTITLEMENTS = HutchWidgetExtension/HutchWidgetExtension.entitlements; CODE_SIGN_STYLE = Automatic; - CURRENT_PROJECT_VERSION = 11; + CURRENT_PROJECT_VERSION = 12; DEVELOPMENT_TEAM = ZCNAX3VL9D; GENERATE_INFOPLIST_FILE = NO; INFOPLIST_FILE = HutchWidgetExtension/Info.plist; @@ -615,7 +615,7 @@ "@executable_path/Frameworks", "@executable_path/../../Frameworks", ); - MARKETING_VERSION = 2.4.1; + MARKETING_VERSION = 2.4.2; PRODUCT_BUNDLE_IDENTIFIER = net.cleberg.Hutch.HutchWidgetExtension; PRODUCT_NAME = "$(TARGET_NAME)"; SKIP_INSTALL = YES; diff --git a/Hutch/Views/Repositories/ReadmeView.swift b/Hutch/Views/Repositories/ReadmeView.swift index dea41f3..27cec55 100644 --- a/Hutch/Views/Repositories/ReadmeView.swift +++ b/Hutch/Views/Repositories/ReadmeView.swift @@ -283,8 +283,15 @@ nonisolated func markdownToHTML(_ text: String, imageURLResolver: ((String) -> S let lines = normalizedText.split(separator: "\n", omittingEmptySubsequences: false).map(String.init) var html = "" var inCodeBlock = false - var inList = false + var codeBlockInListItem = false + var codeBlockLines: [String] = [] + var listType: MarkupListType? + var inBlockquote = false + var pendingListItemBreak = false + var currentListItemLines: [String] = [] + var currentListItemBlocks: [String] = [] var paragraph: [String] = [] + var tableRows: [[String]] = [] func flushParagraph() { if !paragraph.isEmpty { @@ -296,72 +303,262 @@ nonisolated func markdownToHTML(_ text: String, imageURLResolver: ((String) -> S } } + func flushListItem() { + guard !currentListItemLines.isEmpty || !currentListItemBlocks.isEmpty else { return } + let itemContent = currentListItemLines + .map { $0.trimmingCharacters(in: .whitespaces) } + .joined(separator: " ") + + if currentListItemBlocks.isEmpty { + html += "<li>" + renderTaskListItem( + itemContent, + inlineRenderer: { processInline($0, imageURLResolver: imageURLResolver) } + ) + "</li>\n" + } else { + if !itemContent.isEmpty { + currentListItemBlocks.append( + "<p>" + renderTaskListItem( + itemContent, + inlineRenderer: { processInline($0, imageURLResolver: imageURLResolver) } + ) + "</p>" + ) + } + html += "<li>" + currentListItemBlocks.joined(separator: "\n") + "</li>\n" + } + currentListItemLines = [] + currentListItemBlocks = [] + } + + func flushListItemParagraphIntoBlocks() { + guard !currentListItemLines.isEmpty else { return } + let itemContent = currentListItemLines + .map { $0.trimmingCharacters(in: .whitespaces) } + .joined(separator: " ") + currentListItemBlocks.append( + "<p>" + renderTaskListItem( + itemContent, + inlineRenderer: { processInline($0, imageURLResolver: imageURLResolver) } + ) + "</p>" + ) + currentListItemLines = [] + } + + func flushCodeBlock() { + let content = codeBlockLines.joined(separator: "\n") + let blockHTML = "<pre><code>" + content + "</code></pre>\n" + if codeBlockInListItem { + currentListItemBlocks.append(blockHTML) + } else { + html += blockHTML + } + codeBlockLines = [] + codeBlockInListItem = false + } + func closeList() { - if inList { + flushListItem() + switch listType { + case .unordered: html += "</ul>\n" - inList = false + case .ordered: + html += "</ol>\n" + case nil: + break + } + listType = nil + } + + func flushTable() { + guard !tableRows.isEmpty else { return } + html += renderHTMLTable( + rows: tableRows, + inlineRenderer: { processInline($0, imageURLResolver: imageURLResolver) } + ) + tableRows = [] + } + + func closeBlockquote() { + if inBlockquote { + flushParagraph() + html += "</blockquote>\n" + inBlockquote = false } } for line in lines { + let trimmed = line.trimmingCharacters(in: .whitespaces) + + if pendingListItemBreak, listType != nil { + if trimmed.isEmpty { + continue + } + if isIndentedContinuationLine(line) || trimmed.hasPrefix("```") { + pendingListItemBreak = false + } else if isMarkdownUnorderedListItem(trimmed) || orderedListItem(in: trimmed) != nil { + flushListItem() + pendingListItemBreak = false + } else { + flushListItem() + closeList() + pendingListItemBreak = false + } + } + + if let rawHTML = sanitizedMarkdownHTMLLine(from: trimmed) { + closeBlockquote() + flushParagraph() + flushTable() + if listType != nil { + flushListItemParagraphIntoBlocks() + currentListItemBlocks.append(rawHTML) + } else { + closeList() + html += rawHTML + "\n" + } + continue + } + // Fenced code blocks - if line.hasPrefix("```") { + if trimmed.hasPrefix("```") { if inCodeBlock { - html += "</code></pre>\n" + flushCodeBlock() inCodeBlock = false } else { + closeBlockquote() flushParagraph() - closeList() - html += "<pre><code>" + flushTable() + codeBlockInListItem = listType != nil && (!currentListItemLines.isEmpty || !currentListItemBlocks.isEmpty) + if !codeBlockInListItem { + closeList() + } else { + flushListItemParagraphIntoBlocks() + } inCodeBlock = true + codeBlockLines = [] } continue } if inCodeBlock { - html += escapeHTML(line) + "\n" + codeBlockLines.append(escapeHTML(line)) continue } + if isTableLine(trimmed) { + closeBlockquote() + flushParagraph() + closeList() + tableRows.append(parseTableRow(trimmed)) + continue + } else { + flushTable() + } + // Headings + if line.hasPrefix("###### ") { + closeBlockquote() + flushParagraph() + closeList() + html += "<h6>" + processInline(String(line.dropFirst(7)), imageURLResolver: imageURLResolver) + "</h6>\n" + continue + } + if line.hasPrefix("##### ") { + closeBlockquote() + flushParagraph() + closeList() + html += "<h5>" + processInline(String(line.dropFirst(6)), imageURLResolver: imageURLResolver) + "</h5>\n" + continue + } + if line.hasPrefix("#### ") { + closeBlockquote() + flushParagraph() + closeList() + html += "<h4>" + processInline(String(line.dropFirst(5)), imageURLResolver: imageURLResolver) + "</h4>\n" + continue + } if line.hasPrefix("### ") { + closeBlockquote() flushParagraph() closeList() html += "<h3>" + processInline(String(line.dropFirst(4)), imageURLResolver: imageURLResolver) + "</h3>\n" continue } if line.hasPrefix("## ") { + closeBlockquote() flushParagraph() closeList() html += "<h2>" + processInline(String(line.dropFirst(3)), imageURLResolver: imageURLResolver) + "</h2>\n" continue } if line.hasPrefix("# ") { + closeBlockquote() flushParagraph() closeList() html += "<h1>" + processInline(String(line.dropFirst(2)), imageURLResolver: imageURLResolver) + "</h1>\n" continue } + if isMarkdownHorizontalRule(trimmed) { + closeBlockquote() + flushParagraph() + closeList() + html += "<hr>\n" + continue + } + + if trimmed.hasPrefix("> ") { + flushTable() + closeList() + if !inBlockquote { + flushParagraph() + html += "<blockquote>\n" + inBlockquote = true + } + paragraph.append(processInline(String(trimmed.dropFirst(2)), imageURLResolver: imageURLResolver)) + continue + } else { + closeBlockquote() + } + // List items - let trimmed = line.trimmingCharacters(in: .whitespaces) if trimmed.hasPrefix("- ") || trimmed.hasPrefix("* ") { flushParagraph() - if !inList { + if listType != .unordered { + closeList() html += "<ul>\n" - inList = true + listType = .unordered } - html += "<li>" + renderTaskListItem( - String(trimmed.dropFirst(2)), - inlineRenderer: { processInline($0, imageURLResolver: imageURLResolver) } - ) + "</li>\n" + flushListItem() + currentListItemLines = [String(trimmed.dropFirst(2))] + continue + } + if let orderedItem = orderedListItem(in: trimmed) { + flushParagraph() + if listType != .ordered { + closeList() + html += "<ol>\n" + listType = .ordered + } + flushListItem() + currentListItemLines = [orderedItem] + continue + } + + if listType != nil && isIndentedContinuationLine(line) { + currentListItemLines.append(trimmed) continue } // Blank line if trimmed.isEmpty { - flushParagraph() - closeList() + if inBlockquote { + closeBlockquote() + } else if listType != nil, !currentListItemLines.isEmpty || !currentListItemBlocks.isEmpty { + pendingListItemBreak = true + } else { + flushParagraph() + closeList() + } continue } @@ -371,21 +568,33 @@ nonisolated func markdownToHTML(_ text: String, imageURLResolver: ((String) -> S // Flush remaining state if inCodeBlock { - html += "</code></pre>\n" + flushCodeBlock() } + closeBlockquote() flushParagraph() + flushTable() closeList() return html } nonisolated func processInline(_ text: String, imageURLResolver: ((String) -> String?)? = nil) -> String { - var result = escapeHTML(text) + var protectedFragments: [String: String] = [:] + var result = protectMatches( + in: text, + pattern: #"</?[A-Za-z][^>]*?>"#, + protectedFragments: &protectedFragments + ) { match, nsText in + let rawTag = nsText.substring(with: match.range) + return sanitizedMarkdownHTMLTag(rawTag) ?? escapeHTML(rawTag) + } + + result = escapeHTML(result) // Images:  result = replaceMatches(in: result, pattern: #"!\[([^\]]*)\]\(([^)]+)\)"#) { match, nsText in let alt = nsText.substring(with: match.range(at: 1)) - let source = nsText.substring(with: match.range(at: 2)) + let source = decodeHTMLEntities(nsText.substring(with: match.range(at: 2))) let resolvedSource = imageURLResolver?(source) ?? source guard let sanitizedSource = sanitizedReadmeImageURLString(resolvedSource) else { return escapeHTML(alt) @@ -395,12 +604,30 @@ nonisolated func processInline(_ text: String, imageURLResolver: ((String) -> St // Links: [text](url) result = replaceMatches(in: result, pattern: #"\[([^\]]+)\]\(([^)]+)\)"#) { match, nsText in let label = nsText.substring(with: match.range(at: 1)) - let rawURL = nsText.substring(with: match.range(at: 2)) + let rawURL = decodeHTMLEntities(nsText.substring(with: match.range(at: 2))) guard let sanitizedURL = sanitizedReadmeLinkURLString(rawURL) else { return label } return #"<a href="\#(sanitizedURL)">\#(label)</a>"# } + // Plain email autolinks + result = replaceMatches( + in: result, + pattern: #"(?i)(?<![\w.%+\-])([A-Z0-9._%+\-]+@[A-Z0-9.\-]+\.[A-Z]{2,})(?![\w\-])"# + ) { match, nsText in + guard !isInsideHTMLTag(nsText, range: match.range) else { + return nsText.substring(with: match.range) + } + let email = nsText.substring(with: match.range(at: 1)) + let href = escapeHTMLAttribute("mailto:\(email)") + return #"<a href="\#(href)">\#(email)</a>"# + } + // Strikethrough: ~~text~~ + result = result.replacingOccurrences( + of: #"~~(.+?)~~"#, + with: "<del>$1</del>", + options: .regularExpression + ) // Bold: **text** result = result.replacingOccurrences( of: #"\*\*(.+?)\*\*"#, @@ -409,7 +636,13 @@ nonisolated func processInline(_ text: String, imageURLResolver: ((String) -> St ) // Italic: *text* result = result.replacingOccurrences( - of: #"\*(.+?)\*"#, + of: #"(?<!\*)\*(?!\*)(.+?)(?<!\*)\*(?!\*)"#, + with: "<em>$1</em>", + options: .regularExpression + ) + // Italic: _text_ + result = result.replacingOccurrences( + of: #"(?<!\w)_(.+?)_(?!\w)"#, with: "<em>$1</em>", options: .regularExpression ) @@ -420,6 +653,10 @@ nonisolated func processInline(_ text: String, imageURLResolver: ((String) -> St options: .regularExpression ) + for (token, fragment) in protectedFragments { + result = result.replacingOccurrences(of: token, with: fragment) + } + return result } @@ -429,12 +666,39 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String let normalizedText = text .replacingOccurrences(of: "\r\n", with: "\n") .replacingOccurrences(of: "\r", with: "\n") - let lines = normalizedText.split(separator: "\n", omittingEmptySubsequences: false).map(String.init) + let rawLines = normalizedText.split(separator: "\n", omittingEmptySubsequences: false).map(String.init) + var title: String? + var author: String? + var date: String? + let lines = rawLines.filter { line in + let trimmed = line.trimmingCharacters(in: .whitespaces) + guard let keywordMatch = trimmed.firstMatch(of: /^#\+([A-Za-z]+):\s*(.*)$/) else { + return true + } + let keyword = String(keywordMatch.1).lowercased() + let value = String(keywordMatch.2).trimmingCharacters(in: .whitespaces) + switch keyword { + case "title": + title = value + return false + case "author": + author = value + return false + case "date": + date = value + return false + default: + return true + } + } var html = "" var listType: OrgListType? var inQuoteBlock = false var inPropertyDrawer = false var srcLanguage: String? + var inExampleBlock = false + var inCenterBlock = false + var currentListItemLines: [String] = [] var paragraph: [String] = [] var tableRows: [[String]] = [] var propertyRows: [(String, String)] = [] @@ -449,7 +713,20 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String } } + func flushListItem() { + guard !currentListItemLines.isEmpty else { return } + let content = currentListItemLines + .map { $0.trimmingCharacters(in: .whitespaces) } + .joined(separator: " ") + html += "<li>" + renderTaskListItem( + content, + inlineRenderer: { processOrgInline($0, imageURLResolver: imageURLResolver) } + ) + "</li>\n" + currentListItemLines = [] + } + func closeList() { + flushListItem() switch listType { case .unordered: html += "</ul>\n" @@ -463,34 +740,10 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String func flushTable() { guard !tableRows.isEmpty else { return } - let hasHeaderSeparator = tableRows.count > 1 && tableRows[1].allSatisfy(isOrgTableSeparatorCell) - let headerRow = tableRows.first ?? [] - let bodyRows: [[String]] - - html += "<table>\n" - if hasHeaderSeparator { - html += "<thead><tr>" - for cell in headerRow { - html += "<th>" + processOrgInline(cell, imageURLResolver: imageURLResolver) + "</th>" - } - html += "</tr></thead>\n<tbody>\n" - bodyRows = Array(tableRows.dropFirst(2)) - } else { - bodyRows = tableRows - } - - for row in bodyRows { - html += "<tr>" - for cell in row { - html += "<td>" + processOrgInline(cell, imageURLResolver: imageURLResolver) + "</td>" - } - html += "</tr>\n" - } - - if hasHeaderSeparator { - html += "</tbody>\n" - } - html += "</table>\n" + html += renderHTMLTable( + rows: tableRows, + inlineRenderer: { processOrgInline($0, imageURLResolver: imageURLResolver) } + ) tableRows = [] } @@ -520,6 +773,21 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String } } + func closeExampleBlock() { + if inExampleBlock { + html += "</code></pre>\n" + inExampleBlock = false + } + } + + func closeCenterBlock() { + if inCenterBlock { + flushParagraph() + html += "</div>\n" + inCenterBlock = false + } + } + func flushBlockState() { flushParagraph() closeList() @@ -527,6 +795,20 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String flushPropertyDrawer() } + if title != nil || author != nil || date != nil { + html += "<div class=\"org-metadata\">\n" + if let title { + html += "<h1 class=\"org-title\">" + escapeHTML(title) + "</h1>\n" + } + if let author { + html += "<p class=\"org-author\">" + escapeHTML(author) + "</p>\n" + } + if let date { + html += "<p class=\"org-date\">" + escapeHTML(date) + "</p>\n" + } + html += "</div>\n" + } + for line in lines { let trimmed = line.trimmingCharacters(in: .whitespaces) @@ -539,11 +821,35 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String continue } + if inExampleBlock { + if trimmed.lowercased() == "#+end_example" { + closeExampleBlock() + } else { + html += escapeHTML(line) + "\n" + } + continue + } + if inQuoteBlock, trimmed.lowercased() == "#+end_quote" { closeQuoteBlock() continue } + if inCenterBlock { + if trimmed.lowercased() == "#+end_center" { + closeCenterBlock() + } else if trimmed.isEmpty { + flushParagraph() + } else { + paragraph.append(line) + } + continue + } + + if trimmed == "#" || trimmed.hasPrefix("# ") { + continue + } + if trimmed.lowercased().hasPrefix("#+begin_src") { closeQuoteBlock() flushBlockState() @@ -559,6 +865,14 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String continue } + if trimmed.lowercased() == "#+begin_example" { + closeQuoteBlock() + flushBlockState() + html += "<pre><code>" + inExampleBlock = true + continue + } + if trimmed.lowercased() == "#+begin_quote" { flushBlockState() html += "<blockquote>\n" @@ -566,6 +880,14 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String continue } + if trimmed.lowercased() == "#+begin_center" { + closeQuoteBlock() + flushBlockState() + html += "<div style=\"text-align:center\">\n" + inCenterBlock = true + continue + } + if trimmed == ":PROPERTIES:" { closeQuoteBlock() flushBlockState() @@ -592,18 +914,25 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String } } - if isOrgTableLine(trimmed) { + if isTableLine(trimmed) { closeQuoteBlock() flushParagraph() closeList() - tableRows.append(parseOrgTableRow(trimmed)) + tableRows.append(parseTableRow(trimmed)) continue } else { flushTable() } + if isOrgHorizontalRule(trimmed) { + closeQuoteBlock() + flushBlockState() + html += "<hr>\n" + continue + } + // Org headings: * heading, ** heading, *** heading - if let match = trimmed.firstMatch(of: /^(\*{1,3})\s+(.+)$/) { + if let match = trimmed.firstMatch(of: /^(\*{1,6})\s+(.+)$/) { closeQuoteBlock() flushBlockState() let level = match.1.count @@ -621,10 +950,8 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String html += "<ul>\n" listType = .unordered } - html += "<li>" + renderTaskListItem( - String(trimmed.dropFirst(2)), - inlineRenderer: { processOrgInline($0, imageURLResolver: imageURLResolver) } - ) + "</li>\n" + flushListItem() + currentListItemLines = [String(trimmed.dropFirst(2))] continue } @@ -636,10 +963,13 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String html += "<ol>\n" listType = .ordered } - html += "<li>" + renderTaskListItem( - orderedItem, - inlineRenderer: { processOrgInline($0, imageURLResolver: imageURLResolver) } - ) + "</li>\n" + flushListItem() + currentListItemLines = [orderedItem] + continue + } + + if listType != nil && isIndentedContinuationLine(line) { + currentListItemLines.append(trimmed) continue } @@ -658,6 +988,8 @@ nonisolated func orgToHTML(_ text: String, imageURLResolver: ((String) -> String } closeSourceBlock() + closeExampleBlock() + closeCenterBlock() closeQuoteBlock() flushBlockState() @@ -720,6 +1052,22 @@ nonisolated private func processOrgInline(_ text: String, imageURLResolver: ((St } return "<code>\(codeText)</code>" } + result = protectMatches( + in: result, + pattern: #"(?<!\S)\+(.+?)\+(?=\s|$|[.,;:!?])"#, + protectedFragments: &protectedFragments + ) { match, nsText in + let value = nsText.substring(with: match.range(at: 1)) + return "<del>\(value)</del>" + } + result = protectMatches( + in: result, + pattern: #"(?<!\S)_(.+?)_(?=\s|$|[.,;:!?])"#, + protectedFragments: &protectedFragments + ) { match, nsText in + let value = nsText.substring(with: match.range(at: 1)) + return "<u>\(value)</u>" + } // Bold: *text* result = result.replacingOccurrences( @@ -733,6 +1081,17 @@ nonisolated private func processOrgInline(_ text: String, imageURLResolver: ((St with: "<em>$1</em>", options: .regularExpression ) + result = replaceMatches( + in: result, + pattern: #"(?i)(?<![\w.%+\-])([A-Z0-9._%+\-]+@[A-Z0-9.\-]+\.[A-Z]{2,})(?![\w\-])"# + ) { match, nsText in + guard !isInsideHTMLTag(nsText, range: match.range) else { + return nsText.substring(with: match.range) + } + let email = nsText.substring(with: match.range(at: 1)) + let href = escapeHTMLAttribute("mailto:\(email)") + return #"<a href="\#(href)">\#(email)</a>"# + } for (token, fragment) in protectedFragments { result = result.replacingOccurrences(of: token, with: fragment) @@ -805,11 +1164,11 @@ nonisolated private func sanitizeReadmeURLString( return escapeHTMLAttribute(sanitizedURL) } -nonisolated private func isOrgTableLine(_ line: String) -> Bool { +nonisolated private func isTableLine(_ line: String) -> Bool { line.hasPrefix("|") && line.hasSuffix("|") } -nonisolated private func parseOrgTableRow(_ line: String) -> [String] { +nonisolated private func parseTableRow(_ line: String) -> [String] { line .split(separator: "|", omittingEmptySubsequences: false) .dropFirst() @@ -817,21 +1176,163 @@ nonisolated private func parseOrgTableRow(_ line: String) -> [String] { .map { String($0).trimmingCharacters(in: .whitespaces) } } -nonisolated private func isOrgTableSeparatorCell(_ cell: String) -> Bool { +nonisolated private func isTableSeparatorCell(_ cell: String) -> Bool { let trimmed = cell.trimmingCharacters(in: .whitespaces) return !trimmed.isEmpty && trimmed.allSatisfy { $0 == "-" || $0 == "+" } } -private enum OrgListType { +nonisolated private func renderHTMLTable( + rows: [[String]], + inlineRenderer: (String) -> String +) -> String { + guard !rows.isEmpty else { return "" } + let hasHeaderSeparator = rows.count > 1 && rows[1].allSatisfy(isTableSeparatorCell) + let headerRow = rows.first ?? [] + let bodyRows = hasHeaderSeparator ? Array(rows.dropFirst(2)) : rows + var html = "<table>\n" + + if hasHeaderSeparator { + html += "<thead><tr>" + for cell in headerRow { + html += "<th>" + inlineRenderer(cell) + "</th>" + } + html += "</tr></thead>\n" + } + + html += "<tbody>\n" + for row in bodyRows { + html += "<tr>" + for cell in row { + html += "<td>" + inlineRenderer(cell) + "</td>" + } + html += "</tr>\n" + } + html += "</tbody>\n" + html += "</table>\n" + return html +} + +private enum MarkupListType: Equatable { case unordered case ordered } +private typealias OrgListType = MarkupListType + nonisolated private func orderedListItem(in line: String) -> String? { guard let match = line.firstMatch(of: /^(\d+)\.\s+(.+)$/) else { return nil } return String(match.2) } +nonisolated private func isMarkdownHorizontalRule(_ line: String) -> Bool { + matchesRegex(line, pattern: #"^\s*([*\-_])(?:\s*\1){2,}\s*$"#) +} + +nonisolated private func isOrgHorizontalRule(_ line: String) -> Bool { + matchesRegex(line, pattern: #"^\s*-{5,}\s*$"#) +} + +nonisolated private func matchesRegex(_ text: String, pattern: String) -> Bool { + guard let regex = try? NSRegularExpression(pattern: pattern) else { return false } + let range = NSRange(location: 0, length: (text as NSString).length) + return regex.firstMatch(in: text, range: range) != nil +} + +nonisolated private func isInsideHTMLTag(_ text: NSString, range: NSRange) -> Bool { + guard range.location != NSNotFound else { return false } + let prefix = text.substring(to: range.location) + guard let lastOpen = prefix.lastIndex(of: "<") else { return false } + guard let lastClose = prefix.lastIndex(of: ">") else { return true } + return lastOpen > lastClose +} + +nonisolated private func isIndentedContinuationLine(_ line: String) -> Bool { + guard !line.trimmingCharacters(in: .whitespaces).isEmpty else { return false } + guard let first = line.first else { return false } + return first == " " || first == "\t" +} + +nonisolated private func isMarkdownUnorderedListItem(_ line: String) -> Bool { + line.hasPrefix("- ") || line.hasPrefix("* ") +} + +nonisolated private func decodeHTMLEntities(_ text: String) -> String { + text + .replacingOccurrences(of: "&", with: "&") + .replacingOccurrences(of: """, with: "\"") + .replacingOccurrences(of: "'", with: "'") + .replacingOccurrences(of: "<", with: "<") + .replacingOccurrences(of: ">", with: ">") +} + +nonisolated private func sanitizedMarkdownHTMLLine(from line: String) -> String? { + guard line.hasPrefix("<"), line.hasSuffix(">") else { return nil } + return sanitizedMarkdownHTMLTag(line) +} + +nonisolated private func sanitizedMarkdownHTMLTag(_ rawTag: String) -> String? { + let trimmed = rawTag.trimmingCharacters(in: .whitespacesAndNewlines) + guard trimmed.hasPrefix("<"), trimmed.hasSuffix(">") else { return nil } + guard !trimmed.lowercased().hasPrefix("<!--") else { return nil } + + let selfClosing = trimmed.hasSuffix("/>") + let contentStart = trimmed.index(after: trimmed.startIndex) + let contentEnd = trimmed.index(trimmed.endIndex, offsetBy: selfClosing ? -2 : -1) + let inner = trimmed[contentStart..<contentEnd].trimmingCharacters(in: .whitespacesAndNewlines) + let isClosing = inner.hasPrefix("/") + let body = isClosing ? inner.dropFirst().trimmingCharacters(in: .whitespacesAndNewlines) : inner + let parts = body.split(separator: " ", maxSplits: 1, omittingEmptySubsequences: true) + guard let rawName = parts.first else { return nil } + let tagName = rawName.lowercased() + let allowedTags: Set<String> = [ + "a", "abbr", "b", "blockquote", "br", "code", "del", "div", "em", + "hr", "i", "img", "li", "ol", "p", "pre", "span", "strong", "sub", + "sup", "u", "ul" + ] + guard allowedTags.contains(tagName) else { return nil } + + if isClosing { + return "</\(tagName)>" + } + + let attributePortion = parts.count > 1 ? String(parts[1]) : "" + let attributes = parseHTMLAttributes(attributePortion) + var renderedAttributes: [String] = [] + + for (name, value) in attributes { + switch (tagName, name.lowercased()) { + case ("a", "href"): + if let sanitized = sanitizedReadmeLinkURLString(decodeHTMLEntities(value)) { + renderedAttributes.append(#"href="\#(sanitized)""#) + } + case ("img", "src"): + if let sanitized = sanitizedReadmeImageURLString(decodeHTMLEntities(value)) { + renderedAttributes.append(#"src="\#(sanitized)""#) + } + case ("img", "alt"), (_, "title"), (_, "class"): + renderedAttributes.append(#"\#(name)="\#(escapeHTMLAttribute(value))""#) + default: + continue + } + } + + let suffix = selfClosing || tagName == "br" || tagName == "hr" || tagName == "img" ? " /" : "" + let attributeText = renderedAttributes.isEmpty ? "" : " " + renderedAttributes.joined(separator: " ") + return "<\(tagName)\(attributeText)\(suffix)>" +} + +nonisolated private func parseHTMLAttributes(_ text: String) -> [(String, String)] { + guard let regex = try? NSRegularExpression(pattern: #"([A-Za-z_:][A-Za-z0-9:._-]*)\s*=\s*"([^"]*)""#) else { + return [] + } + let nsText = text as NSString + return regex.matches(in: text, range: NSRange(location: 0, length: nsText.length)).map { match in + let name = nsText.substring(with: match.range(at: 1)) + let value = nsText.substring(with: match.range(at: 2)) + return (name, value) + } +} + nonisolated private func protectMatches( in text: String, pattern: String, @@ -843,7 +1344,7 @@ nonisolated private func protectMatches( let matches = regex.matches(in: result, range: NSRange(location: 0, length: (result as NSString).length)) for match in matches.reversed() { - let token = "__ORG_PROTECTED_\(protectedFragments.count)__" + let token = "ZZPROTECTED\(protectedFragments.count)ZZ" let nsText = result as NSString protectedFragments[token] = transform(match, nsText) result = nsText.replacingCharacters(in: match.range, with: token) @@ -877,8 +1378,9 @@ nonisolated private func makeOrgImageHTML( ) -> String? { guard isRenderableImageSource(source) else { return nil } let resolvedSource = imageURLResolver?(source) ?? source + guard let sanitizedSource = sanitizedReadmeImageURLString(resolvedSource) else { return nil } let altText = escapeHTMLAttribute(alt ?? "") - return #"<img src="\#(resolvedSource)" alt="\#(altText)">"# + return #"<img src="\#(sanitizedSource)" alt="\#(altText)">"# } nonisolated private func isRenderableImageSource(_ source: String) -> Bool { @@ -911,7 +1413,19 @@ nonisolated func resolveRepositoryAssetURL( } guard !relativePath.isEmpty else { return nil } - return "https://git.sr.ht/\(owner)/\(repositoryName)/blob/HEAD/\(relativePath)" + var components = URLComponents() + components.scheme = "https" + components.host = "git.sr.ht" + let encodedOwner = owner.addingPercentEncoding(withAllowedCharacters: .urlPathAllowed) ?? owner + let encodedRepository = repositoryName.addingPercentEncoding(withAllowedCharacters: .urlPathAllowed) ?? repositoryName + let encodedRelativePath = relativePath + .split(separator: "/", omittingEmptySubsequences: false) + .map { segment in + String(segment).addingPercentEncoding(withAllowedCharacters: .urlPathAllowed) ?? String(segment) + } + .joined(separator: "/") + components.percentEncodedPath = "/\(encodedOwner)/\(encodedRepository)/blob/HEAD/\(encodedRelativePath)" + return components.string } nonisolated private func normalizeRepositoryPath(_ path: String) -> String { @@ -1099,6 +1613,53 @@ private struct HTMLWebViewRepresentable: UIViewRepresentable { a { color: \(linkColor); } table { border-collapse: collapse; width: 100%; } td, th { border: 1px solid #ccc; padding: 4px 8px; } + blockquote { + border-left: 3px solid rgba(128, 128, 128, 0.5); + margin: 0.5em 0; + padding: 0.25em 0 0.25em 1em; + color: inherit; + opacity: 0.85; + } + hr { + border: none; + border-top: 1px solid rgba(128, 128, 128, 0.35); + margin: 1em 0; + } + table { + border-collapse: collapse; + width: 100%; + margin: 0.75em 0; + font-size: 0.95em; + } + th { + background: rgba(128, 128, 128, 0.15); + font-weight: 600; + text-align: left; + } + td, th { + border: 1px solid rgba(128, 128, 128, 0.3); + padding: 6px 10px; + } + dl.org-properties { + margin: 0.5em 0; + display: grid; + grid-template-columns: max-content 1fr; + gap: 2px 12px; + } + dt { + font-weight: 600; + font-family: ui-monospace, Menlo, monospace; + font-size: 0.9em; + } + dd { margin: 0; } + .org-metadata { margin-bottom: 1em; } + .org-title { margin: 0 0 0.25em; } + .org-author, .org-date { + margin: 0; + color: rgba(128, 128, 128, 0.85); + font-size: 0.9em; + } + del { opacity: 0.7; } </style> </head> <body>\(html)</body> diff --git a/HutchTests/ReadmeViewTests.swift b/HutchTests/ReadmeViewTests.swift index 9c3c019..7fbca86 100644 --- a/HutchTests/ReadmeViewTests.swift +++ b/HutchTests/ReadmeViewTests.swift @@ -27,3 +27,214 @@ struct ReadmeViewTests { #expect(sanitizedReadmeLinkURLString("#readme") == "#readme") } } + +struct MarkdownRenderingTests { + + @Test + func markdownOrderedList() { + let html = markdownToHTML("1. First\n2. Second\n3. Third") + + #expect(html.contains("<ol>")) + #expect(html.contains("<li>First</li>")) + #expect(html.contains("<li>Third</li>")) + #expect(html.contains("</ol>")) + } + + @Test + func markdownBlockquote() { + let html = markdownToHTML("> This is a quote") + + #expect(html.contains("<blockquote>")) + #expect(html.contains("This is a quote")) + #expect(html.contains("</blockquote>")) + } + + @Test + func markdownTable() { + let input = "| A | B |\n|---|---|\n| 1 | 2 |" + let html = markdownToHTML(input) + + #expect(html.contains("<table>")) + #expect(html.contains("<th>")) + #expect(html.contains("<td>")) + } + + @Test + func markdownHorizontalRule() { + let html = markdownToHTML("---") + + #expect(html.contains("<hr>")) + } + + @Test + func markdownDeepHeadings() { + let html = markdownToHTML("#### Level 4\n##### Level 5\n###### Level 6") + + #expect(html.contains("<h4>")) + #expect(html.contains("<h5>")) + #expect(html.contains("<h6>")) + } + + @Test + func markdownStrikethrough() { + let html = markdownToHTML("~~deleted~~") + + #expect(html.contains("<del>deleted</del>")) + } + + @Test + func markdownInlineCodeEscaping() { + let html = processInline("`<b>`") + + #expect(html.contains("<code>")) + #expect(html.contains("<b>")) + #expect(!html.contains("<b>")) + } + + @Test + func markdownWrappedBulletNormalizesLines() { + let html = markdownToHTML("- First line\n continues here") + + #expect(html.contains("<li>First line continues here</li>")) + } + + @Test + func markdownPlainEmailAutolinks() { + let html = processInline("Contact [email protected]") + + #expect(html.contains(#"href="mailto:[email protected]""#)) + #expect(html.contains(">[email protected]</a>")) + } + + @Test + func markdownListContinuesAfterCodeBlock() { + let html = markdownToHTML(""" + 1. Clone the repository: + ```sh + git clone https://git.sr.ht/~ccleberg/Hutch + ``` + 2. Open the project in Xcode. + """) + + #expect(html.contains("<ol>")) + #expect(html.contains("<pre><code>")) + #expect(html.contains("<li><p>Clone the repository:</p>")) + #expect(html.contains("<li>Open the project in Xcode.</li>")) + #expect(html.contains("</ol>")) + } + + @Test + func markdownListContinuesAfterBlankLineIndentedCodeBlock() { + let html = markdownToHTML(""" + 1. Clone the repository: + + ```sh + git clone https://git.sr.ht/~ccleberg/Hutch + ``` + + 2. Open the project in Xcode. + """) + + #expect(html.contains("<li><p>Clone the repository:</p>")) + #expect(html.contains("<pre><code>git clone https://git.sr.ht/~ccleberg/Hutch</code></pre>")) + #expect(html.contains("<li>Open the project in Xcode.</li>")) + #expect(!html.contains("<ol>\n<li>Open the project in Xcode.</li>\n</ol>\n<ol>")) + #expect(html.firstRange(of: "<p>Clone the repository:</p>")!.lowerBound < html.firstRange(of: "<pre><code>git clone https://git.sr.ht/~ccleberg/Hutch</code></pre>")!.lowerBound) + } + + @Test + func markdownImageQueryStringPreservesAmpersands() { + let html = processInline("") + + #expect(html.contains("metric=security_rating")) + #expect(!html.contains("amp;metric")) + #expect(html.contains("<img")) + } + + @Test + func markdownAllowsSafeInlineHTML() { + let html = processInline(#"<strong>Bold</strong> <a href="https://example.com">Link</a>"#) + + #expect(html.contains("<strong>Bold</strong>")) + #expect(html.contains(#"<a href="https://example.com">Link</a>"#)) + } + + @Test + func markdownProtectedTokensDoNotLeak() { + let html = processInline(#"<a href="https://example.com">Back to top</a> `yoshi [ARG] <FILE>`"#) + + #expect(!html.contains("ZZPROTECTED")) + #expect(html.contains(#"<a href="https://example.com">Back to top</a>"#)) + #expect(html.contains("<code>yoshi [ARG] <FILE></code>")) + } +} + +struct OrgRenderingTests { + + @Test + func orgDeepHeading() { + let html = orgToHTML("**** Level 4 Heading") + + #expect(html.contains("<h4>")) + } + + @Test + func orgCommentLinesIgnored() { + let html = orgToHTML("# This is a comment\nNormal text") + + #expect(!html.contains("This is a comment")) + #expect(html.contains("Normal text")) + } + + @Test + func orgStrikethrough() { + let html = orgToHTML("+deleted text+") + + #expect(html.contains("<del>")) + } + + @Test + func orgExampleBlock() { + let html = orgToHTML("#+begin_example\nhello world\n#+end_example") + + #expect(html.contains("<pre><code>")) + #expect(html.contains("hello world")) + } + + @Test + func orgTitleKeyword() { + let html = orgToHTML("#+TITLE: My Document\nBody text") + + #expect(html.contains("org-title")) + #expect(html.contains("My Document")) + } + + @Test + func orgItalicDoesNotMatchURLPaths() { + let html = orgToHTML("[[https://example.com/path/to/file]]") + + #expect(!html.contains("<em>")) + } + + @Test + func orgWrappedBulletNormalizesLines() { + let html = orgToHTML("- First line\n continues here") + + #expect(html.contains("<li>First line continues here</li>")) + } +} + +struct RepositoryAssetURLTests { + + @Test + func repositoryAssetURLPercentEncodesImagePaths() { + let url = resolveRepositoryAssetURL( + "images/My Logo.png", + owner: "~ccleberg", + repositoryName: "Hutch", + readmePath: "README.md" + ) + + #expect(url == "https://git.sr.ht/~ccleberg/Hutch/blob/HEAD/images/My%20Logo.png") + } +} |