Parser for collecting data from Rozetka product ratings
It is necessary to create a parser for collecting data from Rozetka and filling it into a table.
Sample table at the link:
The table is needed for analyzing the product ratings in the Rozetka search results. The table includes 4 products as an example. In fact, there can be any number of products.
Columns A, B are mandatory to fill, their values are filled in by a person. Columns C, D, E are filled in by the parser. Columns F, G are optional and can be of any number or may not exist at all; they are created by a person with any headers and values. Columns H and further to the right are filled in by the parser.
The column "Search Query" - the query entered in the search.
The column "Product Name" - the name of the product at the link.
The column "Product Category" - the name of the final product category (example in the screenshot).
The column "Category Yes/No". If the search query leads to a product category, for example, the link https://bt.rozetka.com.ua/ua/252133/c252133/, then the value is "yes"; if it does not lead, and the query remains a search query, for example, the link https://rozetka.com.ua/ua/search/?text=%D1%84%D0%BE%D1%82%D0%BE%D0%B5%D0%BF%D1%96%D0%BB%D1%8F%D1%82%D0%BE%D1%80§ion_id=252133&redirected=1, then the value is "no".
The column "Ranking". For each search query, a place in the ranking of the results is filled in the table. It is filled in green if the place is from 1 to 5, yellow if the place is from 6 to 20, orange if the place is below 21 but on the first page, red if the product is on the second page or lower. If the product is not found, a dash is placed and colored red. In the product results, there are advertising spots, marked with the word "Advertisement" (example in the screenshot); if the product has such a spot, "ad" is added to the number.
The column "Advertisement". For each search query, a place in the advertising results on the page is filled from left to right; if the product is not present, a dash is indicated, in parentheses the number of advertising offers on the page. It is filled in green if 1 product is in the top 10, yellow if below 10th place but present in the advertisement, red if there is no product in the advertising results. The advertising results are in the product cards (example in the screenshot), the parser looks in the first card of the organic results. If there is no product in the first card of the advertising results, it looks in the second; if there is none in the second, it looks in the third; if there is none in the third, it sets the value to "-" and does not mark it with color.
The column "Top Sales". If the card indicates "Top Sales" (example in the screenshot) - the value is "yes" in green color; if not, the value is "no" in yellow color. If the card has an advertising spot in the organic results, "Top Sales" may not be shown in the results but only in the card, then it is necessary to check in the card (example in the screenshot).
The parser collects data in the table every day for all products entered in the table, at the time specified by a person during setup. The history of parsing must be preserved, meaning new data is added to the table while old data remains in it. If the table runs out, a second identical one is created, and the parser fills it. All old tables are preserved.
In the parsing settings, a person specifies the number of pages of results for data collection. When adding links to products, there should be an option to add a list of links. Products are arranged in the table from top to bottom alphabetically by the first letter of the name.
The parser must operate in Incognito mode so that history and cookies do not interfere with the parser seeing the correct product results.
Access must be provided through a web browser via login and password. The user registers themselves with any login and password, and the admin confirms their registration, after which the user gains access. If necessary, the admin deletes the user, and the user's access is revoked. The number of users is unlimited.
After the project is completed and accepted, ongoing support is needed to ensure the parser operates without errors.
-
Hello, I worked on a project for parsing a marketplace, where we collected data on hundreds of products daily ✅ We gathered over two thousand rows in a table.
Do I understand correctly that automatic processing of all states of advertising positions in Rozetka's output is needed?
I suggest we get in touch, I will consult you for free on the technical side and we will create a development plan + I will tell you about my team!