Server: Editing notes with image insertion basically works

src/calibre/srv/content.py

@@ -4,11 +4,14 @@
 __license__ = 'GPL v3'
 __copyright__ = '2015, Kovid Goyal <kovid at kovidgoyal.net>'
 
+import base64
 import errno
 import os
+import re
 from contextlib import suppress
 from functools import partial
 from io import BytesIO
+from json import load as load_json_file
 from threading import Lock
 
 from calibre import fit_image, guess_type, sanitize_file_name
@@ -22,7 +25,7 @@ from calibre.ebooks.metadata import authors_to_string
 from calibre.ebooks.metadata.meta import set_metadata
 from calibre.ebooks.metadata.opf2 import metadata_to_opf
 from calibre.library.save_to_disk import find_plugboard
-from calibre.srv.errors import BookNotFound, HTTPNotFound
+from calibre.srv.errors import BookNotFound, HTTPBadRequest, HTTPNotFound
 from calibre.srv.routes import endpoint, json
 from calibre.srv.utils import get_db, get_use_roman, http_date
 from calibre.utils.config_base import tweaks
@@ -355,7 +358,14 @@ def get(ctx, rd, what, book_id, library_id):
                 raise HTTPNotFound(f'No {what.lower()} format for the book {book_id!r}')
 
 
-@endpoint('/get-note/{field}/{item_id}/{library_id=None}')
+def resource_hash_to_url(ctx, scheme, digest, library_id):
+    kw = {'scheme': scheme, 'digest': digest}
+    if library_id:
+        kw['library_id'] = library_id
+    return ctx.url_for('/get-note-resource', **kw)
+
+
+@endpoint('/get-note/{field}/{item_id}/{library_id=None}', types={'item_id': int})
 def get_note(ctx, rd, field, item_id, library_id):
     db = get_db(ctx, rd, library_id)
     if db is None:
@@ -374,10 +384,7 @@ def get_note(ctx, rd, field, item_id, library_id):
         pat = re.compile(rf'{RESOURCE_URL_SCHEME}://({{}})'.format('|'.join(map(r, resources))))
         def sub(m):
             s, d = m.group(1).split('/', 1)
-            kw = {'scheme': s, 'digest': d}
+            return resource_hash_to_url(ctx, s, d, library_id)
-            if library_id:
-                kw['library_id'] = library_id
-            return ctx.url_for('/get-note-resource', **kw)
         note_data['doc'] = pat.sub(sub, html)
     rd.outheaders['Content-Type'] = 'text/html; charset=UTF-8'
     rd.outheaders['Last-Modified'] = http_date(note_data['mtime'])
@@ -398,3 +405,50 @@ def get_note_resource(ctx, rd, scheme, digest, library_id):
         fname_for_content_disposition(name), fname_for_content_disposition(name, as_encoded_unicode=True))
     rd.outheaders['Last-Modified'] = http_date(d['mtime'])
     return d['data']
+
+
+@endpoint('/set-note/{field}/{item_id}/{library_id=None}', needs_db_write=True, methods={'POST'}, types={'item_id': int})
+def set_note(ctx, rd, field, item_id, library_id):
+    db = get_db(ctx, rd, library_id)
+    if db is None:
+        raise HTTPNotFound(f'Library {library_id} not found')
+    try:
+        data = load_json_file(rd.request_body_file)
+        if not isinstance(data, dict):
+            raise Exception('note data must be a dict')
+        html, searchable_text, images = data['html'], data['searchable_text'], data['images']
+    except Exception as err:
+        raise HTTPBadRequest(f'Invalid query: {err}')
+    srv_replacements = {}
+    db_replacements = {}
+    resources = []
+    res_pat = re.compile(r'get-note-resource/([a-zA-Z0-9]+)/([a-zA-Z0-9]+)')
+    for key, img in images.items():
+        try:
+            is_new_image = img['data'].startswith('data:')
+            if is_new_image:
+                d = img['data'].encode('ascii')
+                idx = d.index(b',')
+                d = memoryview(d)[idx:]
+                img_data = base64.standard_b64decode(d)
+                fname = img['filename']
+            else:
+                m = res_pat.search(img['data'])
+                scheme, digest = m.group(1), m.group(2)
+                resources.append(f'{scheme}:{digest}')
+        except Exception as err:
+            raise HTTPBadRequest(f'Invalid query: {err}')
+        if is_new_image:
+            chash = db.add_notes_resource(img_data, fname)
+            scheme, digest = chash.split(':', 1)
+            resources.append(chash)
+        srv_replacements[key] = resource_hash_to_url(ctx, scheme, digest, library_id)
+        db_replacements[key] = f'{RESOURCE_URL_SCHEME}://{scheme}/{digest}'
+    db_html = srv_html = html
+    if db_replacements:
+        db_html = re.sub('|'.join(map(re.escape, db_replacements)), lambda m: db_replacements[m.group()], html)
+    if srv_replacements:
+        srv_html = re.sub('|'.join(map(re.escape, srv_replacements)), lambda m: srv_replacements[m.group()], html)
+    db.set_notes_for(field, item_id, db_html, searchable_text, resources, True)
+    rd.outheaders['Content-Type'] = 'text/html; charset=UTF-8'
+    return srv_html

src/pyj/book_list/comments_editor.pyj

@@ -419,21 +419,16 @@ class CommentsEditorBoss:
         document.body.lastChild.innerHTML = data.html
 
     def get_html(self, data):
-        c = document.body.lastChild
+        c = document.body.lastChild.cloneNode(True)
         images = v'{}'
-        for img in c.getElementByTagName('img'):
+        if data.extract_images:
-            if img.src and img.src.startsWith('data:'):
+            for img in c.getElementsByTagName('img'):
-                key = short_uuid4()
+                if img.src:
-                images[key] = {'data': img.src, 'filename': img.dataset.filename}
+                    key = short_uuid4()
-                img.src = key
+                    images[key] = {'data': img.src, 'filename': img.dataset.filename}
-                v'delete img.dataset.filename'
+                    img.src = key
-        self.comm.send_message('html', html=c.innerHTML, extra_data={'images': images})
+                    v'delete img.dataset.filename'
-        for img in c.getElementByTagName('img'):
+        self.comm.send_message('html', html=c.innerHTML, extra_data={'images': images, 'text': c.innerText})
-            if img.src:
-                d = images[img.src]
-                if d:
-                    img.src = d.data
-                    img.dataset.filename = d.filename
 
     def exec_command(self, data):
         document.execCommand(data.name, False, data.value)
@@ -469,6 +464,7 @@ class Editor:
         self.id = iframe.id
         self.ready = False
         self.pending_set_html = None
+        self.pending_get_html = v'[]'
         self.get_html_callbacks = v'[]'
         self.iframe_obj = iframe
 
@@ -496,8 +492,9 @@ class Editor:
         if self.pending_set_html is not None:
             self.set_html(self.pending_set_html)
             self.pending_set_html = None
-        if self.get_html_callbacks.length:
+        if self.pending_get_html.length > 0:
-            self.get_html(self.get_html_callback)
+            for x in self.pending_get_html:
+                self.get_html(x.proceed, x.extract_images)
 
     def set_html(self, html):
         if not self.ready:
@@ -506,15 +503,17 @@ class Editor:
         rgba = get_color_as_rgba('window-foreground')
         self.iframe_wrapper.send_message('set_html', html=html, color_scheme=color_scheme(), color=f'rgba({rgba[0]},{rgba[1]},{rgba[2]},{rgba[3]})')
 
-    def get_html(self, proceed):
+    def get_html(self, proceed, extract_images):
         self.get_html_callbacks.push(proceed)
         if self.ready:
-            self.iframe_wrapper.send_message('get_html')
+            self.iframe_wrapper.send_message('get_html', extract_images=v'!!extract_images')
+        else:
+            self.pending_get_html.push({'proceed': proceed, 'extract_images': v'!!extract_images'})
 
     def on_html_received(self, data):
         if self.get_html_callbacks.length:
             for f in self.get_html_callbacks:
-                f(data.html, data.img_urls)
+                f(data.html, data.extra_data)
             self.get_html_callbacks = v'[]'
 
     def exec_command(self, name, value=None):
@@ -607,11 +606,11 @@ def set_comments_html(container, html):
     editor.set_html(html or '')
 
 
-def get_comments_html(container, proceed):
+def get_comments_html(container, proceed, extract_images):
     iframe = container.querySelector('iframe')
     eid = iframe.getAttribute('id')
     editor = registry[eid]
-    editor.get_html(proceed)
+    editor.get_html(proceed, extract_images)
 
 
 def develop(container):

src/pyj/book_list/show_note.pyj

@@ -13,6 +13,7 @@ from book_list.router import back, home, show_note
 from book_list.top_bar import add_button, create_top_bar
 from book_list.ui import set_panel_handler
 from gettext import gettext as _
+from modals import ajax_send_progress_dialog, error_dialog
 from utils import parse_url_params, safe_set_inner_html, sandboxed_html
 
 
@@ -144,11 +145,29 @@ def save(container_id):
     c = document.getElementById(container_id)
     if not c:
         return
-    get_comments_html(c, def(html, extra_data):
+    q = parse_url_params()
-        print(html)
+    url = 'set-note/' + encodeURIComponent(q.field) + '/' + encodeURIComponent(q.item_id)
-        console.log(extra_data)
+    if q.library_id:
-        print('TODO: Implement saving')
+        url += '/' + encodeURIComponent(q.library_id)
-    )
+    close_action = get_close_action()[0]
+
+    def on_complete(end_type, xhr, evt):
+        nonlocal current_note_markup
+        if end_type is 'abort':
+            close_action()
+            return
+        if end_type is not 'load':
+            error_dialog(_('Failed to send notes to server'), _(
+                'Updating the notes for: {} failed.').format(q.item), xhr.error_html)
+            return
+        q.html = xhr.responseText
+        current_note_markup = q
+        close_action()
+
+    def on_got_html(html, extra_data):
+        ajax_send_progress_dialog(url, {'html': html, 'searchable_text': extra_data.text, 'images': extra_data.images}, on_complete, _('Sending note data to calibre server...'))
+
+    get_comments_html(c, on_got_html, True)
 
 
 def init_edit(container_id):