feat: add issue view with body cleaning and truncation (task 0003)

Add `issue view <n>` as the first detail command, introducing the
content-cleaning and truncation machinery (src/body.ts) that later issue
and PR slices reuse.

- Default output: number, title, state, author, created, body (truncated
  at 500), plus comment_count; --comments expands every comment (bodies
  truncated at 800); --full suppresses all truncation.
- cleanBody runs only when a body exceeds its limit: normalizes Gitea
  issue/PR URLs on the detected host to Issue#N/PR#N, strips image embeds
  and long URLs, and collapses quoted blocks.
- Type guard: a PR number fails with VALIDATION_ERROR and a `pr view <n>`
  hint, detected via the fetched object's pull_request field.
- renderDetail joins the detail entity, optional sub-blocks, and help.

src/body.ts

@@ -0,0 +1,90 @@
+/**
+ * Body cleaning and truncation, shared by every command that renders an issue,
+ * pull request, or comment body. Cleaning is deliberately applied *only* when a
+ * body exceeds its truncation limit, so short bodies pass through byte-for-byte
+ * (see the gitea-axi spec, "Content truncation").
+ */
+
+export const BODY_TRUNCATE_LIMIT = 500;
+export const COMMENT_TRUNCATE_LIMIT = 800;
+
+function escapeForRegex(value: string): string {
+  return value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+}
+
+/** The two Gitea URL path segments that carry a reference number. */
+type RefKind = "issues" | "pulls";
+
+function issueRef(kind: RefKind, number: string): string {
+  return kind === "pulls" ? `PR#${number}` : `Issue#${number}`;
+}
+
+/**
+ * Normalize Gitea issue/PR URLs on the given host, strip image embeds and long
+ * URLs, and collapse email-style quoted blocks. Transforms run in a fixed order
+ * so that later, coarser rules never consume text an earlier rule owns.
+ */
+export function cleanBody(text: string, host: string): string {
+  const escapedHost = escapeForRegex(host);
+  // owner/repo/(issues|pulls)/N, with an optional #fragment or ?query tail.
+  const urlCore = `https?://${escapedHost}/[^/\\s)]+/[^/\\s)]+/(issues|pulls)/(\\d+)`;
+
+  let result = text;
+
+  // 1. Gitea issue/PR URLs inside markdown links collapse to the bare ref.
+  result = result.replace(
+    new RegExp(`\\[[^\\]]*\\]\\(${urlCore}[^)]*\\)`, "g"),
+    (_match, kind: RefKind, number: string) => issueRef(kind, number),
+  );
+
+  // 2. Bare Gitea issue/PR URLs collapse to the ref as well.
+  result = result.replace(
+    new RegExp(`${urlCore}(?:[#?][^\\s)]*)?`, "g"),
+    (_match, kind: RefKind, number: string) => issueRef(kind, number),
+  );
+
+  // 3. Markdown image embeds become a compact placeholder.
+  result = result.replace(
+    /!\[([^\]]*)\]\([^)]*\)/g,
+    (_match, alt: string) => (alt.trim() ? `[image: ${alt.trim()}]` : "[image]"),
+  );
+
+  // 4. Markdown links wrapping a long URL (>80 chars) keep only their label.
+  result = result.replace(
+    /\[([^\]]*)\]\(([^)]+)\)/g,
+    (match, label: string, url: string) => (url.length > 80 ? `[${label}]` : match),
+  );
+
+  // 5. Standalone long URLs (>100 chars) are removed entirely.
+  result = result.replace(/https?:\/\/[^\s)]+/g, (match) =>
+    match.length > 100 ? "[long URL removed]" : match,
+  );
+
+  // 6. Email-style quoted blocks of 3+ consecutive `>` lines collapse to a note.
+  result = result.replace(
+    /(?:^[ \t]*>.*(?:\r?\n|$)){3,}/gm,
+    "[quoted text removed]\n",
+  );
+
+  return result;
+}
+
+/**
+ * Return a body for display. Bodies within `maxLen` are returned untouched. When
+ * over the limit, cleaning is applied: if the cleaned body now fits it is
+ * returned with a "cleaned" note; otherwise it is truncated with the inline
+ * "truncated" hint. `N` in both notes is the original body length, since that is
+ * what `--full` would reveal.
+ */
+export function truncateBody(body: string, maxLen: number, host: string): string {
+  if (body.length <= maxLen) {
+    return body;
+  }
+  const cleaned = cleanBody(body, host);
+  if (cleaned.length <= maxLen) {
+    // Reaching here means cleaning shortened an over-limit body to fit, so it
+    // necessarily changed the content — hence the unconditional note.
+    return `${cleaned}\n(cleaned, ${body.length} chars original - use --full to see original)`;
+  }
+  return `${cleaned.slice(0, maxLen)}\n... (truncated, ${body.length} chars total - use --full to see complete body)`;
+}