Elbuz adds a supplier website as a price list source: the wizard links it to a ready-made parser or creates a universal parser from the URL and tries to pick parsing rules.
When you need this:
- The website has a ready-made parser — rules are written, no selectors needed.
- No ready-made parser — the universal one finds the sitemap, picks selectors and collects 300 trial products.
- Only a few sections needed — their URLs on the “Links” tab limit crawling.
- Product URLs are in a sitemap or file — links are imported from XML or CSV.
Where to start: “List of price lists” → connection wizard → “Site parser” tile.
What the wizard can do for a parser
- Link a source to a ready-made parser — Rozetka, PROM, Hotline, Kaspi and other marketplaces, no selectors to configure.
- Create a universal parser — for any website by URL, with auto configuration and a trial run.
- Set the starting URLs for crawling — manually, from an sitemap or from a file.
How to choose the parser type in the wizard
You add the source with the same connection wizard as a file price list. At the “Who” step, choose the contact; at the “From where” step, choose the “Site parser” tile labeled “scraping site pages”. Two blocks appear below the tiles, and they are mutually exclusive: picking a ready-made parser clears the universal parser field and vice versa. If you pick neither, the wizard says “Pick a ready-made parser from the list or enter a website link”.
1
Site parser tile2
Ready-made parsers3
Universal parser, website link4
Auto configuration switch5
Load without opening product page- The Site parser tile takes data from website pages rather than from a file.
- Ready-made parsers — marketplaces and stores that Elbuz supports out of the box.
- Universal parser — any other website, entered as a link in the field.
- Auto configuration (start loading first 300 products) is on by default, so collection starts by itself after the wizard.
- Load products and prices without opening the product page.
- The website is on the ready-made parser listRozetka, PROM, Hotline, Kaspi, eBay, Amazon and other marketplaces — the parsing rules are already written
- The website is not on the listUniversal parser: you enter the URL, and selectors are picked automatically or manually
A ready-made parser from the list
The “Ready-made parsers” block has the hint “Stores and marketplaces we maintain ourselves — pick one and configure it.” The list currently has fourteen modules, including Rozetka and Rozetka (), PROM, Hotline and Hotline (API), Kaspi, eMAG, Black Red White, public.cy, eBay, Amazon, B&H Photo Video and Прайс навигатор (Price Navigator).
Choose a module and click “Next”: the source is linked to it, and the settings window opens right away. No selectors to write: a ready-made module has no “Import settings” and “Testing” tabs at all, only “List of links”, “Seller list” and “Additional”. Some modules with their own API also get a “My products” tab in front — that is price monitoring, covered in another article.
The “Seller list” tab only matters for aggregator marketplaces where different stores sell the same product: it holds the seller allowlist and blocklist, a markup for each seller and the choice of price source. For a regular website it is empty.
A ready-made module’s settings are shared: there is no per-source copy, every source links to the same system parser record. With two Rozetka sources (yours and a competitor’s), changing a checkbox in one changes the other. If a record has several sources, the window says so above the tabs: “These parser settings are shared: N sources use them. A change here affects all of them.” For different settings on one website, add a universal as the second source.
A universal parser for any website
The “Universal parser” block has the hint “Any website: enter a link, CSS selectors are configured manually.” The “Website link” field shows the example “for example https://site.com”; if you leave out the scheme, the program adds https:// itself. Below are two toggles, both on by default:
- “Auto configuration (start loading first 300 products)” — the program checks whether the website has a and tries to pick selectors itself;
- “Load products and prices without opening the product page” — names and prices are taken straight from the section page, product pages are not opened. Faster, but nothing comes from the product page: no description, no specifications.
After “Next”, the program does a few things for you. It downloads the website home page and extracts the organization details from its markup — logo, address, description, phone, email; all of this is added to the contact card, but only to empty fields, so your data is not overwritten. It also takes the website favicon, which becomes the parser icon.
It creates 64 empty parsing rules across four operations. Exactly one of them is filled: the h1 selector for the product name on the product page. Then, if auto configuration is left on, it looks for sitemap.xml, and if there is none, for a Sitemap: line in robots.txt, and turns on .
- Action or result
- Decision: the path depends on the answer
With the wizard’s default settings, collection starts on its own — and not only when a sitemap is found. If the second toggle, “Load products and prices without opening the product page”, is on, the sitemap is not searched for at all and is treated as found, so the wizard opens the “Import data from the site” window with autostart in any case. The only way to avoid this is to turn off “Auto configuration” before “Next”: the settings window then opens and collection waits for your click.
The first after auto configuration is deliberately short: no more than 50 section links and no more than 300 saved products, after which the task queue is cleared and collection stops on its own. Pagination pages are not added in the first run either — only the first page of each section is taken.
This lets you look at the result before crawling covers the whole website. The limit applies exactly once: when the run finishes, it is lifted, and the next run has no cap.
How to set the URLs crawling starts from
The “Links” tab is the list of starting URLs. An empty list is not an error: for a universal parser it means “start from the home page and find the sections yourself”. But if you only need three sections out of thirty, or the website has no catalog at all, enter the URLs here.
Grid columns: “Status”, “Link”, “This is a link to a product”, “This is a link to a product list”, “Name”, “Manufacturer SKU”, “Delivery cost”, “Note”. For a universal parser, only one checkbox matters here — “This is a link to a product”: it sends the URL straight to the “Product card” operation. A URL without this checkbox goes to the “List of product links” operation, that is, it is treated as a section page. The “This is a link to a product list” column does not affect crawling — the CSV import sets it for its own use.
1
Add (plus) button2
This is a link to a product3
List-link column4
Additional operations- Add — paste the addresses of sections and products here, several at once.
- This is a link to a product — the address leads straight to a product card.
- Link-to-list column — the address leads to a category from which product links are collected; in this window it has a narrow header without a title (the column is titled differently than in the script).
- Additional operations — the burger menu with refresh, clear filter and delete-all-links items.
The “+” button opens the “Add link” dialog with a “List of links” field and an “All product links” checkbox. You paste URLs in bulk: the program detects the separator itself — line break or semicolon, whichever occurs more. Duplicates are dropped. If a line contains a tab, the program checks which of the two parts holds the URL; the other part becomes the name.
The second way to fill the list is the “Configure loading links from Sitemap (XML)” accordion on the “Additional” tab. Enter the “Link to sitemap (XML file)”, optionally the text patterns for ignoring or, conversely, selecting URLs, and click “Import links”. The feed then shows the lines “Loading sitemap:”, “Sitemap links loaded:” and “Site page links loaded:”.
By the way, auto configuration enters four patterns in the ignore field — category, categori, article and /filter/ — to keep articles and filter pages out of the list, but only if the sitemap was found at the direct URL site/sitemap.xml. When the URL came from robots.txt, the field stays empty and you enter the patterns yourself.
The third way is the neighboring accordion for a CSV file: column numbers for SKU, name, link and the “this is a link to a product” flag, column and row separators, the row to start from, and the “Clear the link list before uploading a file” checkbox.
The “Additional Operations” menu on this tab has three items: “Refresh list”, “Clear filter” and “Delete all links”. The last one removes the whole list in one go, and there is no undo. If you typed the URLs by hand, export them first — there will be nowhere to restore them from.
Scenarios
You need a few website sections, not the whole catalog
On the “Links” tab, use the “+” button to paste the URLs of the sections you need, and leave “This is a link to a product” unchecked. Each such URL goes to the “List of product links” operation: the program collects the section’s products and pagination pages according to that group’s rules.
What won’t happen: the “List of product category links” operation does not run while the list has at least one URL — the program will not look for other website sections. And if “Delete products before loading” is on, products collected earlier from other sections disappear from the price list at the start of the run.
Don’t start collection right after the wizard
In the “Universal parser” block, turn off “Auto configuration (start loading first 300 products)” before “Next”. The settings window opens, and collection waits for your click — you can set up the parsing rules first.
What won’t happen: the program will not pick selectors; only h1 for the product name stays filled.
Product URLs are already in the sitemap
On the “Additional” tab, in the “Configure loading links from Sitemap (XML)” accordion, enter the “Link to sitemap (XML file)”, optionally the text patterns for ignoring or selecting URLs, and click “Import links”. URLs from the sitemap go into the “Links” list.
What won’t happen: if the sitemap field stays filled, the sitemap is downloaded again on every run, not just once.
FAQ
Why did collection start on its own after the wizard when I didn’t click anything?
That is how the “Auto configuration” toggle works; it is on by default. Together with “Load products and prices without opening the product page”, it opens the collection window with autostart. To prevent this, turn the toggle off before “Next” (about auto configuration).
How many products will the first run collect?
After auto configuration — no more than 50 section links and 300 products, first section pages only. The limit applies once: the next run has no cap.
The wizard says “Pick a ready-made parser from the list or enter a website link”. What’s wrong?
The “Site parser” tile is selected, but no module is chosen and no URL is entered. Choose one of the two: the options are mutually exclusive (about the wizard).
I changed the settings of one source, and another source changed too. Why?
Both use a ready-made module: no copy of the settings is made for your price list. If you need different settings for the same website, add a universal parser — it has its own record (about ready-made parsers).
The wizard takes the organization details from the website. Will my contact data be overwritten?
No. The logo, address, description, phone and email are added to the contact card only in empty fields.
Where to find it and who has access
- Menu
- Directories → Processing price lists → List of price lists
- Wizard
- source connection wizard, the “From where” step → the “Site parser” tile
- Starting URLs
- source settings window, the “Links” tab (“List of links” for ready-made parsers)
- Access right
- the “Price lists” window
- Have ready
- the contact you are adding the source for, and the website URL
Related topics
- For the overall picture of collecting data from websites — Site parsing.
- If the source is added and you need to specify where the name and price are on the page — How to set up page parsing rules.
- If the rules are ready and it’s time to collect — How to start collecting from a website.
- For other source types: file, link, email, FTP — Connecting a supplier source.
- If you connect competitor websites to track their prices — what it gives your store.
