etassia 0.1.0

data/lib/etassia/parser/grammars/mindmap.rb

@@ -0,0 +1,220 @@
+# frozen_string_literal: true
+
+require_relative "common"
+
+module Etassia
+  module Parser
+    module Grammars
+      # Parslet grammar for Mindmap diagrams
+      class Mindmap < Common
+        rule(:diagram) do
+          space? >>
+            header >>
+            node_line.repeat(0).as(:nodes) >>
+            space?
+        end
+
+        # Mermaid does not require whitespace between the "mindmap" keyword
+        # and the root node: any content left on the header's own line
+        # becomes the root, handled below by node_line falling through to
+        # node_with_content with a zero-length indent.
+        rule(:header) do
+          str("mindmap") >> match["a-zA-Z0-9_"].absent? >> eof.absent? >>
+            header_tail
+        end
+
+        # A run of separators after the keyword is only ever discarded when
+        # nothing but a comment or the line end follows it: mermaid's own
+        # lexer never treats that whitespace as a token, so there is no
+        # indent to preserve. When something else follows on the same line
+        # (an inline root), that whitespace IS the root's SPACELIST token in
+        # mermaid's grammar and sets mindmapDb's baseLevel for every later
+        # line's indent to be measured against -- so here we consume nothing
+        # and leave it for node_line's own indent capture.
+        rule(:header_tail) do
+          (match[' \t'].repeat >> comment >> (newline | eof)) |
+            (match[' \t'].repeat >> (newline | eof)) |
+            str("")
+        end
+
+        rule(:node_line) do
+          empty_line | comment_line | node_with_content
+        end
+
+        # A whole-line "%%" comment is a blank line to mermaid's lexer
+        # (SPACELINE), so it never becomes a node.
+        rule(:comment_line) do
+          space? >> comment >> (newline | eof)
+        end
+
+        rule(:empty_line) do
+          space? >> newline
+        end
+
+        # A shaped node's own rule stops at its shape's closing delimiter, so
+        # a same-line trailing "%%" comment -- valid anywhere in mermaid's
+        # lexer -- is tolerated here, once, for every SHAPED node type,
+        # rather than inside each shape rule. node_plain has no closing
+        # delimiter: its content rule (`match('[^\r\n:]').repeat(1)`, below)
+        # already greedily swallows a trailing "%%..." into :content before
+        # this `comment.maybe` ever runs, so a plain node's trailing comment
+        # is NOT stripped. Pre-existing, not introduced by this change;
+        # left as-is here, out of scope for this fix.
+        rule(:node_with_content) do
+          match[' \t'].repeat.as(:indent) >>
+            node >>
+            space? >>
+            comment.maybe >>
+            (newline | eof)
+        end
+
+        rule(:node) do
+          node_with_icon |
+            node_with_class |
+            node_with_shape |
+            node_plain
+        end
+
+        # Node with icon: ::icon(fa fa-book)
+        rule(:node_with_icon) do
+          str("::icon(") >>
+            match("[^)]").repeat(1).as(:icon) >>
+            str(")")
+        end
+
+        # Node with class: :::className
+        rule(:node_with_class) do
+          str(":::") >>
+            match('[^\r\n]').repeat(1).as(:classes)
+        end
+
+        # Node with shape (optional identifier prefix). round_shape must be
+        # tried last: circle_shape's "((" is a strict prefix of a lone "(",
+        # so round_shape would otherwise swallow the first paren of a circle
+        # node and leave its own close paren unconsumed.
+        rule(:node_with_shape) do
+          match["a-zA-Z0-9_"].repeat >>
+            (
+              circle_shape | bang_shape | cloud_shape | hexagon_shape |
+                square_shape | round_shape
+            )
+        end
+
+        # ((text)) - circle
+        rule(:circle_shape) do
+          str("((") >>
+            match("[^)]").repeat(1).as(:content) >>
+            str("))") >>
+            str("").as(:shape_circle)
+        end
+
+        # ))text(( - bang
+        rule(:bang_shape) do
+          str("))") >>
+            match("[^(]").repeat(1).as(:content) >>
+            str("((") >>
+            str("").as(:shape_bang)
+        end
+
+        # )text( - cloud
+        rule(:cloud_shape) do
+          str(")") >>
+            match("[^(]").repeat(1).as(:content) >>
+            str("(") >>
+            str("").as(:shape_cloud)
+        end
+
+        # {{text}} - hexagon
+        rule(:hexagon_shape) do
+          str("{{") >>
+            match("[^}]").repeat(1).as(:content) >>
+            str("}}") >>
+            str("").as(:shape_hexagon)
+        end
+
+        # [text] - square (handles quoted content like ["text with []"])
+        rule(:square_shape) do
+          str("[") >>
+            (
+              # Quoted content: ["text..."]
+              (str('"') >> match('[^"]').repeat(1).as(:content) >> str('"')) |
+              # Regular content without quotes
+              match('[^\]]').repeat(1).as(:content)
+            ) >>
+            str("]") >>
+            str("").as(:shape_square)
+        end
+
+        # (text) - round (rounded rectangle, mermaid's default shape spelled
+        # out explicitly). Content may span multiple physical lines: unlike
+        # node_with_content's own line-at-a-time repeat, nothing inside this
+        # rule stops at a newline, only the closing ")" does -- matching
+        # mermaid's own NODE-state lexer, whose catch-all token for this
+        # shape does not stop at line breaks either. Quoted content (e.g.
+        # `root("a)b")`) mirrors square_shape's own quoted branch: mermaid's
+        # lexer has a dedicated quoted-content state that strips the quotes
+        # and lets a literal ")" through, which the unquoted `[^)]` branch
+        # cannot represent.
+        # Mermaid's own comment-strip regex runs `\s` from JavaScript, which
+        # is far wider than ASCII space/tab -- it also matches a no-break
+        # space and the other Unicode space separators below. Mirrors
+        # Builders::Flowchart::JS_SPACE (flowchart.rb), the same character
+        # set for the identical "indentation before %%" problem.
+        COMMENT_INDENT = '\t\n\v\f\r \u00a0\u1680\u2000-\u200a' \
+                         '\u2028\u2029\u202f\u205f\u3000\ufeff'
+
+        rule(:round_comment_line) do
+          newline >>
+            match[COMMENT_INDENT].repeat >>
+            str("%%") >>
+            str("{").absent? >>
+            match('[^\r\n]').repeat >>
+            newline.maybe
+        end
+
+        rule(:round_unquoted_content) do
+          (round_comment_line | match("[^)]")).repeat(1).as(:content)
+        end
+
+        # A %% comment line embedded inside quoted round content is matched
+        # the SAME way as the unquoted case above, before the closing quote
+        # is ever searched for. Mermaid strips comment lines in a textual
+        # pre-pass that runs before quote lexing, so a literal `"` sitting
+        # inside what will become a stripped comment line does not end the
+        # quoted string in real mermaid -- matching only `[^"]` per
+        # character let such a `"` end the match early, breaking the whole
+        # round_shape alternative and falling back to plain-text nodes.
+        rule(:round_quoted_content) do
+          (round_comment_line | match('[^"]')).repeat(1).as(:content)
+        end
+
+        rule(:round_shape) do
+          str("(") >>
+            (
+              (str('"') >> round_quoted_content >> str('"') >>
+                str("").as(:round_quoted)) |
+              round_unquoted_content
+            ) >>
+            str(")") >>
+            str("").as(:shape_round)
+        end
+
+        # A "(" or "[" with no ")" or "]" after it on the line: mermaid's
+        # lexer opens a shape at the bracket and never finds a closing
+        # token, which is a parse error (mindmap.jison NODE_DSTART).
+        rule(:unterminated_opener) do
+          match('[^\r\n(\[]').repeat >> match("[(\\[]") >>
+            match('[^\r\n)\]]').repeat >> (newline | eof)
+        end
+
+        # Plain text node
+        rule(:node_plain) do
+          unterminated_opener.absent? >>
+            match('[^\r\n:]').repeat(1).as(:content)
+        end
+
+        root(:diagram)
+      end
+    end
+  end
+end

data/lib/etassia/parser/grammars/packet.rb

@@ -0,0 +1,71 @@
+# frozen_string_literal: true
+
+require "parslet"
+require_relative "common"
+
+module Etassia
+  module Parser
+    module Grammars
+      # Grammar for parsing Mermaid packet-beta diagram syntax
+      class Packet < Common
+        rule(:diagram_type) { str("packet-beta") >> space? }
+
+        rule(:text) { match['^\n'].repeat(1) }
+
+        rule(:title_line) do
+          str("title") >> space >> text.as(:title) >> newline
+        end
+
+        # Field definition: 0-10: "Label"
+        rule(:bit_number) do
+          match["0-9"].repeat(1)
+        end
+
+        rule(:bit_range) do
+          bit_number.as(:bit_start) >>
+            (str("-") >> bit_number.as(:bit_end)).maybe
+        end
+
+        rule(:field_label) do
+          str('"') >>
+            ((str("\\") >> any) | match('[^"\\\\]')).repeat(0).as(:label) >>
+            str('"')
+        end
+
+        rule(:field_line) do
+          bit_range >>
+            str(":") >> space? >>
+            field_label >>
+            (newline | eof)
+        end
+
+        rule(:comment_line) do
+          comment >> newline
+        end
+
+        rule(:empty_line) do
+          space? >> newline
+        end
+
+        rule(:indent?) { space? }
+
+        rule(:statement) do
+          indent? >> (
+            title_line |
+            field_line |
+            comment_line |
+            empty_line
+          )
+        end
+
+        rule(:body) { statement.repeat(0) }
+
+        rule(:packet_diagram) do
+          diagram_type >> (newline >> body.as(:statements)).maybe >> eof.maybe
+        end
+
+        root(:packet_diagram)
+      end
+    end
+  end
+end

data/lib/etassia/parser/grammars/pie.rb

@@ -0,0 +1,128 @@
+# frozen_string_literal: true
+
+require_relative "common"
+
+module Etassia
+  module Parser
+    module Grammars
+      # Parslet grammar for pie chart diagrams.
+      #
+      # Handles pie chart syntax including title, data entries with labels
+      # and numeric values, and accessibility features.
+      #
+      # @example Simple pie chart
+      #   pie
+      #     "Apples" : 42.5
+      #     "Oranges" : 30.2
+      #
+      # @example Pie chart with title and showData
+      #   pie title "Sales Report" showData
+      #     "Q1" : 100
+      #     "Q2" : 150
+      class Pie < Common
+        root(:diagram)
+
+        # Main diagram structure
+        rule(:diagram) do
+          ws? >>
+            header >>
+            ws? >>
+            statements.maybe >>
+            ws?
+        end
+
+        # Header: pie [showData] [title <text>]
+        rule(:header) do
+          pie_keyword.as(:header) >>
+            (space.repeat(1) >> show_data_flag).maybe.as(:show_data) >>
+            (space.repeat(1) >> title_declaration).maybe.as(:title) >>
+            ws?
+        end
+
+        # Mermaid's keyword is exactly "pie": "Pie" and "pie chart" are
+        # lexer errors there.
+        rule(:pie_keyword) do
+          str("pie")
+        end
+
+        rule(:show_data_flag) do
+          str("showData").as(:show_data)
+        end
+
+        rule(:title_declaration) do
+          str("title") >> (colon | space.repeat(1)) >>
+            title_text.as(:title)
+        end
+
+        rule(:title_text) do
+          (line_end.absent? >> any).repeat
+        end
+
+        # Statements (can be data entries or accessibility declarations)
+        rule(:statements) do
+          (statement >> ws?).repeat(1)
+        end
+
+        rule(:statement) do
+          acc_title_declaration |
+            acc_descr_declaration |
+            standalone_title_declaration |
+            data_entry |
+            (comment >> line_end)
+        end
+
+        # Title as a standalone statement (not in header)
+        rule(:standalone_title_declaration) do
+          space.repeat >>
+            str("title") >> space.repeat(1) >>
+            title_text.as(:standalone_title) >>
+            line_end
+        end
+
+        # Accessibility title
+        rule(:acc_title_declaration) do
+          str("accTitle") >> space? >> colon >> space? >>
+            (line_end.absent? >> any).repeat.as(:acc_title) >>
+            line_end
+        end
+
+        # Accessibility description (single or multi-line)
+        rule(:acc_descr_declaration) do
+          acc_descr_single_line | acc_descr_multi_line
+        end
+
+        rule(:acc_descr_single_line) do
+          str("accDescr") >> space? >> colon >> space? >>
+            (line_end.absent? >> any).repeat.as(:acc_descr) >>
+            line_end
+        end
+
+        rule(:acc_descr_multi_line) do
+          str("accDescr") >> space? >> lbrace >> ws? >>
+            (rbrace.absent? >> any).repeat.as(:acc_descr) >>
+            ws? >> rbrace >> line_end
+        end
+
+        # Data entry: "label" : value
+        rule(:data_entry) do
+          space.repeat >>
+            label_text.as(:label) >>
+            space.repeat >> colon >>
+            space.repeat >>
+            numeric_value.as(:value) >>
+            line_end.as(:data_entry)
+        end
+
+        # Label can be quoted string or unquoted text
+        rule(:label_text) do
+          quoted_string | single_quoted_string
+        end
+
+        # Numeric value (integer or decimal, can be negative)
+        rule(:numeric_value) do
+          minus.maybe >> (float | integer)
+        end
+      end
+    end
+  end
+end

data/lib/etassia/parser/grammars/quadrant.rb

@@ -0,0 +1,202 @@
+# frozen_string_literal: true
+
+require_relative "common"
+
+module Etassia
+  module Parser
+    module Grammars
+      # Parslet grammar for quadrant chart diagrams.
+      #
+      # Handles quadrant chart syntax including title, axis labels,
+      # quadrant labels, and data points with coordinates and styling.
+      #
+      # @example Simple quadrant chart
+      #   quadrantChart
+      #     title Product Analysis
+      #     x-axis Low Cost --> High Cost
+      #     y-axis Low Value --> High Value
+      #     quadrant-1 Invest
+      #     Product A: [0.3, 0.7]
+      #
+      # @example With styling
+      #   quadrantChart
+      #     Product A: [0.3, 0.7] radius: 10, color: #ff0000
+      class Quadrant < Common
+        root(:diagram)
+
+        # Main diagram structure
+        rule(:diagram) do
+          ws? >>
+            header >>
+            ws? >>
+            statements.maybe >>
+            ws?
+        end
+
+        # Header: quadrantChart
+        rule(:header) do
+          str("quadrantChart").as(:header) >> ws?
+        end
+
+        # Statements (title, axis labels, quadrant labels, points)
+        rule(:statements) do
+          (statement >> ws?).repeat(1)
+        end
+
+        rule(:statement) do
+          title_declaration |
+            x_axis_declaration |
+            y_axis_declaration |
+            quadrant_label_declaration |
+            class_def_declaration |
+            data_point |
+            (comment >> line_end)
+        end
+
+        # Title declaration
+        rule(:title_declaration) do
+          str("title") >> space.repeat(1) >>
+            title_text.as(:title) >>
+            line_end
+        end
+
+        rule(:title_text) do
+          (line_end.absent? >> any).repeat(1)
+        end
+
+        # X-axis: x-axis label1 --> label2
+        rule(:x_axis_declaration) do
+          str("x-axis") >> space.repeat(1) >>
+            axis_label.as(:x_axis_left) >>
+            space? >> str("-->") >> space? >>
+            axis_label.as(:x_axis_right) >>
+            line_end
+        end
+
+        # Y-axis: y-axis label1 --> label2
+        rule(:y_axis_declaration) do
+          str("y-axis") >> space.repeat(1) >>
+            axis_label.as(:y_axis_bottom) >>
+            space? >> str("-->") >> space? >>
+            axis_label.as(:y_axis_top) >>
+            line_end
+        end
+
+        # Axis label (can be quoted or unquoted)
+        rule(:axis_label) do
+          quoted_string | unquoted_axis_label
+        end
+
+        rule(:unquoted_axis_label) do
+          (str("-->").absent? >> line_end.absent? >> any).repeat(1)
+        end
+
+        # Quadrant labels: quadrant-1 through quadrant-4
+        rule(:quadrant_label_declaration) do
+          str("quadrant-") >>
+            match["1-4"].as(:quadrant_number) >>
+            space.repeat(1) >>
+            quadrant_label_text.as(:quadrant_label) >>
+            line_end
+        end
+
+        rule(:quadrant_label_text) do
+          (line_end.absent? >> any).repeat(1)
+        end
+
+        # Class definition: classDef name fill:#color
+        rule(:class_def_declaration) do
+          str("classDef") >> space.repeat(1) >>
+            class_name.as(:class_name) >> space.repeat(1) >>
+            (line_end.absent? >> any).repeat.as(:class_style) >>
+            line_end
+        end
+
+        rule(:class_name) do
+          match["a-zA-Z_"].repeat(1) >> match["a-zA-Z0-9_"].repeat
+        end
+
+        # Data point: label[:::class]: [x, y] [styling...]
+        rule(:data_point) do
+          point_label.as(:label) >>
+            class_reference.maybe.as(:class_ref) >>
+            space? >> colon >> space? >>
+            coordinates.as(:coordinates) >>
+            styling_params.maybe.as(:styling) >>
+            line_end.as(:data_point)
+        end
+
+        # Point label (text before class reference or colon)
+        rule(:point_label) do
+          (str(":::").absent? >> colon.absent? >> any).repeat(1)
+        end
+
+        # Class reference: :::className
+        rule(:class_reference) do
+          str(":::") >> class_name
+        end
+
+        # Coordinates: [x, y]
+        rule(:coordinates) do
+          lbracket >>
+            space? >>
+            float.as(:x) >>
+            space? >> comma >> space? >>
+            float.as(:y) >>
+            space? >>
+            rbracket
+        end
+
+        # Optional styling parameters (flexible whitespace around commas)
+        rule(:styling_params) do
+          (
+            space.repeat(1) >> styling_param >>
+              (space? >> comma >> space? >> styling_param).repeat
+          ).maybe
+        end
+
+        rule(:styling_param) do
+          radius_param |
+            color_param |
+            stroke_color_param |
+            stroke_width_param
+        end
+
+        # radius: value
+        rule(:radius_param) do
+          str("radius") >> space? >> colon >> space? >>
+            (float | integer).as(:radius)
+        end
+
+        # color: #hex
+        rule(:color_param) do
+          str("color") >> space? >> colon >> space? >>
+            color_value.as(:color)
+        end
+
+        # stroke-color: #hex
+        rule(:stroke_color_param) do
+          str("stroke-color") >> space? >> colon >> space? >>
+            color_value.as(:stroke_color)
+        end
+
+        # stroke-width: valuepx
+        rule(:stroke_width_param) do
+          str("stroke-width") >> space? >> colon >> space? >>
+            (float | integer).as(:stroke_width) >>
+            str("px").maybe
+        end
+
+        # Color value (#hex or named color)
+        rule(:color_value) do
+          (str("#") >> match["0-9a-fA-F"].repeat(3, 8)) |
+            match["a-zA-Z"].repeat(1)
+        end
+
+        # Brackets
+        rule(:lbracket) { str("[") }
+        rule(:rbracket) { str("]") }
+      end
+    end
+  end
+end