paliad/pkg/docforge/markdown/importer_test.go

package markdown

import (
	"strings"
	"testing"
)

// Inline-span + block-marker tests, relocated from the docx walker when
// parsing moved here (t-paliad-349 slice 8). parseSpans is the inline
// tokeniser; detectBlockMarker classifies a line.

func TestParseSpans_PlaceholderWithUnderscoresIsLiteral(t *testing.T) {
	// {{project.case_number}} must emit as a single non-italic span
	// containing the full placeholder (the b78a984 fix).
	spans := parseSpans("{{project.case_number}}")
	if len(spans) != 1 {
		t.Fatalf("expected 1 span; got %d (%+v)", len(spans), spans)
	}
	if spans[0].Italic || spans[0].Bold {
		t.Errorf("placeholder must not be italic/bold; got %+v", spans[0])
	}
	if spans[0].Text != "{{project.case_number}}" {
		t.Errorf("placeholder text corrupted: got %q", spans[0].Text)
	}
}

func TestParseSpans_ItalicAroundPlaceholder(t *testing.T) {
	spans := parseSpans("_before_ {{x.y_z}} _after_")
	var saw struct {
		italicBefore bool
		placeholder  bool
		italicAfter  bool
	}
	for _, s := range spans {
		if s.Italic && s.Text == "before" {
			saw.italicBefore = true
		}
		if !s.Italic && !s.Bold && strings.Contains(s.Text, "{{x.y_z}}") {
			saw.placeholder = true
		}
		if s.Italic && s.Text == "after" {
			saw.italicAfter = true
		}
	}
	if !saw.italicBefore || !saw.placeholder || !saw.italicAfter {
		t.Errorf("expected italic/placeholder/italic structure; got %+v", spans)
	}
}

func TestParseSpans_Plain(t *testing.T) {
	spans := parseSpans("hello world")
	if len(spans) != 1 || spans[0].Bold || spans[0].Italic || spans[0].Text != "hello world" {
		t.Errorf("expected single plain span; got %+v", spans)
	}
}

func TestParseSpans_UnderscoreItalic(t *testing.T) {
	spans := parseSpans("_emph_")
	var italicHits int
	for _, s := range spans {
		if s.Italic && s.Text == "emph" {
			italicHits++
		}
	}
	if italicHits != 1 {
		t.Errorf("expected one italic 'emph' span; got %+v", spans)
	}
}

func TestParseSpans_UnderscoreBold(t *testing.T) {
	spans := parseSpans("__strong__")
	var boldHits int
	for _, s := range spans {
		if s.Bold && s.Text == "strong" {
			boldHits++
		}
	}
	if boldHits != 1 {
		t.Errorf("expected one bold 'strong' span; got %+v", spans)
	}
}

func TestDetectBlockMarker(t *testing.T) {
	cases := []struct {
		in   string
		kind string
		want string
		ok   bool
	}{
		{"# A", "heading_1", "A", true},
		{"## B", "heading_2", "B", true},
		{"### C", "heading_3", "C", true},
		{"  # indented", "heading_1", "indented", true}, // up to 3 spaces tolerated
		{"    # too-deep", "", "", false},               // 4 spaces → not a heading
		{"- bullet", "list_bullet", "bullet", true},
		{"* star", "list_bullet", "star", true},
		{"1. one", "list_numbered", "one", true},
		{"42. forty-two", "list_numbered", "forty-two", true},
		{"1) paren", "list_numbered", "paren", true},
		{"1.no-space", "", "", false}, // ordinal needs trailing space
		{"> quote", "blockquote", "quote", true},
		{"plain", "", "", false},
		{"#nospace", "", "", false}, // heading needs space after hash
	}
	for _, tc := range cases {
		t.Run(tc.in, func(t *testing.T) {
			kind, payload, ok := detectBlockMarker(tc.in)
			if ok != tc.ok || kind != tc.kind || payload != tc.want {
				t.Errorf("detectBlockMarker(%q) = (%q,%q,%v); want (%q,%q,%v)", tc.in, kind, payload, ok, tc.kind, tc.want, tc.ok)
			}
		})
	}
}

// TestImport_Document spot-checks the neutral Document the importer
// produces — block kinds, the link-span shape, and placeholder pass-through.
func TestImport_Document(t *testing.T) {
	doc := Import("# Title\n\nBody **bold** and [label](http://x).\n\n- item")
	if len(doc.Blocks) != 3 {
		t.Fatalf("blocks = %d; want 3 (%+v)", len(doc.Blocks), doc.Blocks)
	}
	if doc.Blocks[0].Kind != "heading_1" {
		t.Errorf("block0 kind = %q; want heading_1", doc.Blocks[0].Kind)
	}
	if doc.Blocks[2].Kind != "list_bullet" {
		t.Errorf("block2 kind = %q; want list_bullet", doc.Blocks[2].Kind)
	}
	// The body paragraph carries a link span with Link set + children.
	var sawLink bool
	for _, s := range doc.Blocks[1].Spans {
		if s.Link == "http://x" && len(s.Children) > 0 {
			sawLink = true
		}
	}
	if !sawLink {
		t.Errorf("body block missing link span; got %+v", doc.Blocks[1].Spans)
	}
}

func TestImport_EmptyYieldsOneEmptyParagraph(t *testing.T) {
	doc := Import("")
	if len(doc.Blocks) != 1 || doc.Blocks[0].Kind != "paragraph" || len(doc.Blocks[0].Spans) != 0 {
		t.Errorf("empty import = %+v; want one empty paragraph block", doc.Blocks)
	}
}