easy populate 4, getting a feed to work headache, please help?

[flappingfish]

ok so i have easy populate 4 succesfully installed and i can upload and import a file however it only seems to update the few items in 2 sub categories I had manually listed and skips the rest saying i haven't specified a category?

Even the items that would belong in those 2 sub categories that i've not already listed confusingly. attached a screenshot of the csv file loaded up as you can see "master categories id" has letters in, these letters change also every so often in the same categories as denoted in "v_categories_name_1"



My images have been replaced with what appears to be broken url errors also for the few items it has uploaded as can be seen below...



any help would be appreciated

[mc12345678]

So in a default Zen Cart install, master_categories_id is a number... something has been done to modify the use/storage of that master_categories_id to use the designation(s) listed...

Changing it to letters and importing that data would also impact how a default install of EP4 works and would (have) required modification to support import of this non-numeric designation. Then, there is the issue of assigning product to categories...

For one, the code uses the v_categories_name_X field to detect the existence of the category and then if the category does not exist it creates a new category_id. Also, if the product does not (yet) exist in the database, then the product's master_categories_id is assigned to the category to which the product is first assigned. By default EP4 does not use the master_categories_id when importing. That would require identifying it as a user defined field. This was done basically as a "precaution" because, a conflict could easily be created by assigning both a master_categories_id and a category that is unrelated to that master_categories_id and/or where the master_categories_id does not exist and... and... and...

So... if your ZC store has been modified to use this non-numeric categories_id, then a similar modification would be needed to the EP4 code to create/retrieve/store the non-numeric categories_id as handled by the code.

Now, as for the image(s)... an image of the "image(s)" only helps those here but a very very very small amount. More helpful is to actually share a link to a product/page where this problem is occurring. Second to that would be to provide a copy of the html source code at the point where that image is expected to be shown. Also it would help to explain what has been done to support the images being collected/displayed from the internet instead of from the local server...

[flappingfish]

So in a default Zen Cart install, master_categories_id is a number... something has been done to modify the use/storage of that master_categories_id to use the designation(s) listed...

Changing it to letters and importing that data would also impact how a default install of EP4 works and would (have) required modification to support import of this non-numeric designation. Then, there is the issue of assigning product to categories...

For one, the code uses the v_categories_name_X field to detect the existence of the category and then if the category does not exist it creates a new category_id. Also, if the product does not (yet) exist in the database, then the product's master_categories_id is assigned to the category to which the product is first assigned. By default EP4 does not use the master_categories_id when importing. That would require identifying it as a user defined field. This was done basically as a "precaution" because, a conflict could easily be created by assigning both a master_categories_id and a category that is unrelated to that master_categories_id and/or where the master_categories_id does not exist and... and... and...

So... if your ZC store has been modified to use this non-numeric categories_id, then a similar modification would be needed to the EP4 code to create/retrieve/store the non-numeric categories_id as handled by the code.

Now, as for the image(s)... an image of the "image(s)" only helps those here but a very very very small amount. More helpful is to actually share a link to a product/page where this problem is occurring. Second to that would be to provide a copy of the html source code at the point where that image is expected to be shown. Also it would help to explain what has been done to support the images being collected/displayed from the internet instead of from the local server...

in regards to the pictures, here is an example page where the issue occurs.
https://crazygamer.uk/motherboards-c...dr4-p-710.html

i'm not sure how i would show you the html from that though as thats beyond my knowledge level. I've got autism which helps with this sort of stuff until i get stuck and it just gets me frustrated sadly.

when i did an export from my site to match the headers i had the following fields included....
v_products_model,
v_products_type,
v_products_image,
v_products_name_1,
v_products_description_1,
v_products_url_1,
v_specials_price,
v_specials_date_avail,
v_specials_expires_date,
v_products_price,
v_gtin,
v_products_weight,
v_product_is_call,
v_products_sort_order,
v_products_quantity_order_min,
v_products_quantity_order_units,
v_products_priced_by_attribute,
v_product_is_always_free_shipping,
v_date_avail,
v_date_added,
v_products_quantity,
v_manufacturers_name,
v_categories_name_1,
v_tax_class_title,
v_status,
v_metatags_products_name_status,
v_metatags_title_status,
v_metatags_model_status,
v_metatags_price_status,
v_metatags_title_tagline_status,
v_metatags_title_1,
v_metatags_keywords_1,
v_metatags_description_1

i can customise each field at the suppliers, maybe perhaps i've made a mistake there and should have entered something different? i've attached a screenshot to show how its displayed on the suppliers site for creating the feed.

[mc12345678]

Sorry for the delay in responding...
The above list of fields is typical of a standard installation. Note the column v_categories_name_1. That is the field that allows a full designation of the product to a category/sub-category.

That information needs to be included for a new product to be added to the database. Please take a look at the instructions where there is discussion of the use of the carat (^) as a divider between category designations.

Revisiting the OP of this thread, the image showing the spreadsheet appears to display data that has been exported from the vendor's feed where there is an attempt to capture that data for import. So, there are some of those fields that do not line up with actual ZC fields.

Of those fields shown in the first image, these fields are not likely to be able to be populated by the vendor: v_products_id (numeric and unique to each store and the store's product), v_master_categories_id (again numeric and unique to the category that is found on the branch(es) of categories.

I might actually suggest the field that is currently being called v_products_id probably should be your v_products_model field. I say this because the designation(s) shown in the v_products_id field in the image are likely to be "more unique" than what is shown in the current v_products_model.

Oh, and I forgot to address the other item in the recent post. Keep pushing through, you seem to be doing well at least publicly, even if you're pulling your hair out. Don't know if you knew this about the general public in the forum, but there are also people that are blind that are fully functional in using the forum and ZC, but part of that ability comes from either software that reads text from the screen or a sort of "typewriter", anyways I say that to also suggest that while images kind of work, what is more beneficial is actual text content. It can be posted to the forum "legibly" by encapsulating in [CODE][/CODE] tags.

Back to the category discussion, if there was a category "tree" something like:
Shoes is a main category off of the store.
In the shoes category there are open toed, heels, and flats. Then in the heals category there were two categories: short heels and long heels.

The possible category designation(s) (v_categories_name_1) for a product could be:
Shoes^Open Toed
Shoes^Heels^Short Heals
Shoes^Heels^Lofrng Heals
Shoes^Flats

While EP4 defaults to using the carat for the divider, there are ways to change that to another character if necessary.
I wouldn't expect that there would be many product that would be associated to all four of those test combinations, but it wouldn't be much of a stretch if there was one or two of them.

Related to the images, basically it looks like the software hasn't been modified per se to support using images that are "stored" out in the internet instead of them being captured locally. I say this, because when I visited the above page with the web browser Microsoft Edge, I right clicked on the image text for the main product, then selected the menu option: Inspect. That opened up the html source code and went to the code where the image was being developed...

The html for the image is:

<img src="images/https://pictureserver.co.uk/productimages/200/127097.jpg" alt="ASRock B450 Pro4 AMD Socket AM4 ATX VGA/HDMI/DisplayPort DDR4" title="ASRock B450 Pro4 AMD Socket AM4 ATX VGA/HDMI/DisplayPort DDR4" width="300" height="0">

Notice how the path (src) starts with images/ and then has https:// etc... Well, ZC expects the images to be in the images directory such that whatever is stored in v_products_image is then added to images/... In this situation, that is appending https:// etc to images/ to give content for a link that doesn't really exist or a "broken image".

There's a whole series of modifications (if I remember correctly at the moment) to support using a web address for an image path. I don't recall if it is a "plugin" that helps with this, something in the docs area, or just a forum thread. Its been well over a year since I've looked over the instruction to use web addresses for images.

[flappingfish]

Sorry for the delay in responding...
The above list of fields is typical of a standard installation. Note the column v_categories_name_1. That is the field that allows a full designation of the product to a category/sub-category.

That information needs to be included for a new product to be added to the database. Please take a look at the instructions where there is discussion of the use of the carat (^) as a divider between category designations.

Revisiting the OP of this thread, the image showing the spreadsheet appears to display data that has been exported from the vendor's feed where there is an attempt to capture that data for import. So, there are some of those fields that do not line up with actual ZC fields.

Of those fields shown in the first image, these fields are not likely to be able to be populated by the vendor: v_products_id (numeric and unique to each store and the store's product), v_master_categories_id (again numeric and unique to the category that is found on the branch(es) of categories.

I might actually suggest the field that is currently being called v_products_id probably should be your v_products_model field. I say this because the designation(s) shown in the v_products_id field in the image are likely to be "more unique" than what is shown in the current v_products_model.

Oh, and I forgot to address the other item in the recent post. Keep pushing through, you seem to be doing well at least publicly, even if you're pulling your hair out. Don't know if you knew this about the general public in the forum, but there are also people that are blind that are fully functional in using the forum and ZC, but part of that ability comes from either software that reads text from the screen or a sort of "typewriter", anyways I say that to also suggest that while images kind of work, what is more beneficial is actual text content. It can be posted to the forum "legibly" by encapsulating in [CODE][/CODE] tags.

Back to the category discussion, if there was a category "tree" something like:
Shoes is a main category off of the store.
In the shoes category there are open toed, heels, and flats. Then in the heals category there were two categories: short heels and long heels.

The possible category designation(s) (v_categories_name_1) for a product could be:
Shoes^Open Toed
Shoes^Heels^Short Heals
Shoes^Heels^Lofrng Heals
Shoes^Flats

While EP4 defaults to using the carat for the divider, there are ways to change that to another character if necessary.
I wouldn't expect that there would be many product that would be associated to all four of those test combinations, but it wouldn't be much of a stretch if there was one or two of them.

Related to the images, basically it looks like the software hasn't been modified per se to support using images that are "stored" out in the internet instead of them being captured locally. I say this, because when I visited the above page with the web browser Microsoft Edge, I right clicked on the image text for the main product, then selected the menu option: Inspect. That opened up the html source code and went to the code where the image was being developed...

The html for the image is:

<img src="images/https://pictureserver.co.uk/productimages/200/127097.jpg" alt="ASRock B450 Pro4 AMD Socket AM4 ATX VGA/HDMI/DisplayPort DDR4" title="ASRock B450 Pro4 AMD Socket AM4 ATX VGA/HDMI/DisplayPort DDR4" width="300" height="0">

Notice how the path (src) starts with images/ and then has https:// etc... Well, ZC expects the images to be in the images directory such that whatever is stored in v_products_image is then added to images/... In this situation, that is appending https:// etc to images/ to give content for a link that doesn't really exist or a "broken image".

There's a whole series of modifications (if I remember correctly at the moment) to support using a web address for an image path. I don't recall if it is a "plugin" that helps with this, something in the docs area, or just a forum thread. Its been well over a year since I've looked over the instruction to use web addresses for images.

Thankyou for the advise, i did see the post where you helped the blind person with easy populate feed issues :) it actually helped me to get this far if i remember right :)

so I tried to delete the v_products_model column and renamed the v_products_id column to the "v_categories_model" as suggested (assuming i interpreted your suggestion correctly that is) and it gave me the following result as shown in the image below -there's no messages at the bottom such as product updated or skipped like last time either sadly.



i just edited it using a spreadsheet rather than adjusting it at the suppliers to make it quicker for diagnosing. i think i also tried previously removing the v_products_id column.

just realised before i posted this response also that i stupidly have v_products_model in the feed twice so deleted the one i hadnt edited the header for and uploaded quickly to get exactly the same thing.

In regards to the images ive passed that over to the developers who got stumped as much as i did on the feed part, i hope they can easily rectify that part for me so I can try and get this feed right.

[mc12345678]

Thankyou for the advise, i did see the post where you helped the blind person with easy populate feed issues :) it actually helped me to get this far if i remember right :)

so I tried to delete the v_products_model column and renamed the v_products_id column to the "v_categories_model" as suggested (assuming i interpreted your suggestion correctly that is) and it gave me the following result as shown in the image below -there's no messages at the bottom such as product updated or skipped like last time either sadly.



i just edited it using a spreadsheet rather than adjusting it at the suppliers to make it quicker for diagnosing. i think i also tried previously removing the v_products_id column.

just realised before i posted this response also that i stupidly have v_products_model in the feed twice so deleted the one i hadnt edited the header for and uploaded quickly to get exactly the same thing.

In regards to the images ive passed that over to the developers who got stumped as much as i did on the feed part, i hope they can easily rectify that part for me so I can try and get this feed right.

That message is in relation to the upload (step before importing)... The above description also introduces an opportunity to cause the issue described in the header message (Successfully Uploaded: There are problems with the CSV delimiter(s)).

What that message identifies is that in attempts to review the first few lines of the file, there is some "anomaly" identified that would prevent successfully importing the file.

Unfortunately, it was recently discovered that the process that performs the review/reporting does not play well with files that are stored outside of the admin folder path. I have a modification available to address that; however, haven't packaged it for use. The issue involved there is that when a file path is pushed to the checking software, the file path is "evaluated" to see that from a "security" perspective that it is within the intended folder path and not pointing to somewhere that should not be accessed.

But, if the file is within the admin file path and it is still presenting a CSV delimiter issue, then the CSV file itself is not formatted correctly to support import in the currently available configuration(s).

So... Is your feed file in the catalog side or the admin side? If in the catalog side, I can provide code to address that aspect of the CSV delimiter issue. If in the admin side, then would suggest either re-reviewing the content of the edited CSV file to one obtained from the vendor. Would suggest looking at the overall layout, not the necessarily the details of a particular product... E.g., doesn't really matter in this issue whether the description has the word "double" or "trouble", that doesn't impact the import evaluation, but it is possible that a word with an apostrophe or double quote could be problematic, but is not so automatically.

[flappingfish]

So... Is your feed file in the catalog side or the admin side?

How would i ascertain where my feed file is? code to rectify the issue would be great i'm pretty sure the developers would be eternally grateful

[mc12345678]

How would i ascertain where my feed file is? code to rectify the issue would be great i'm pretty sure the developers would be eternally grateful

Upper right hand corner of the EP4 window: Upload Directory: Does it include the word admin (at the beginning) or not...

I don't (yet) want to cloud the issue here if the file to be imported is *not* in the catalog side.

[flappingfish]

looks like its indeed admin side, ive copy pasted the settings i got on my end. id assume that may also indicate other potential issues :s

Upload Directory: admin/temp/
Primary Key: products_model
Import Language Override: language_code_only
Verbose Feedback: TRUE
Split Records: 2000
Execution Time: 60
Convert Curly Quotes: No Change
Convert Char 0x92: Basic
Enable Products Metatags: 1
Enable Products Music: 0

Custom Products Fields
Product Short Descriptions: FALSE
Product Unit of Measure: FALSE
Product UPC Code: FALSE
Google Product Category: FALSE
Manufacturer's Suggested Retail Price: FALSE
Manufacturer's Advertised Price: FALSE
Group Pricing Per Item: FALSE
Exclusive Products Mod: FALSE
Stock By Attributes Mod: FALSE
CEON URI Rewriter Mod: FALSE
Dual Pricing Mod: FALSE

User Defined Products Fields:
gtin: TRUE
Condition: FALSE

Installed Languages
1-en: English
Default Language: 1-English
Internal Character Encoding: UTF-8
DB Collation: utf8mb4

Database Field Lengths
categories_name:32
manufacturers_name:32
products_model:32
products_name:64
products_url:255
artists_name:32
record_company_name:32
music_genre_name:32


Upload EP File
Http Max Upload File Size: 134217728 bytes (128 Mbytes)

[mc12345678]

looks like its indeed admin side, ive copy pasted the settings i got on my end. id assume that may also indicate other potential issues :s

Upload Directory: admin/temp/
Primary Key: products_model
Import Language Override: language_code_only
Verbose Feedback: TRUE
Split Records: 2000
Execution Time: 60
Convert Curly Quotes: No Change
Convert Char 0x92: Basic
Enable Products Metatags: 1
Enable Products Music: 0

Custom Products Fields
Product Short Descriptions: FALSE
Product Unit of Measure: FALSE
Product UPC Code: FALSE
Google Product Category: FALSE
Manufacturer's Suggested Retail Price: FALSE
Manufacturer's Advertised Price: FALSE
Group Pricing Per Item: FALSE
Exclusive Products Mod: FALSE
Stock By Attributes Mod: FALSE
CEON URI Rewriter Mod: FALSE
Dual Pricing Mod: FALSE

User Defined Products Fields:
gtin: TRUE
Condition: FALSE

Installed Languages
1-en: English
Default Language: 1-English
Internal Character Encoding: UTF-8
DB Collation: utf8mb4

Database Field Lengths
categories_name:32
manufacturers_name:32
products_model:32
products_name:64
products_url:255
artists_name:32
record_company_name:32
music_genre_name:32


Upload EP File
Http Max Upload File Size: 134217728 bytes (128 Mbytes)

So, if all fields of the CSV file are properly formatted and included, then based on the previous post of the fields associated with the export of the database, there will be an issue.

In post #3 at least one of the language dependent fields ends with _1 (a number) which is the language_id of 1. As can be seen above in the quote, the language_id of 1 is English: 1-en: English. But... The Import Language Override is set to be language_code_only (yes default setting at the moment), which for English would be _en, but because the import is only for language_code, then none of that language dependent information will be captured for processing... This was something mentioned multiple times over the last several pages of the easypopulate 4 forum thread. Quite literally, changing that configuration setting from language_code_only to ABSOLUTELY anything else at least will allow the software to process a language dependent field that ends with a number (language_id).

The other thing as discussed earlier, that the user Defined Products Fields: Condition says false, that means that the database table products does not contain the field: Condition, it may contain "condition" or perhaps the field is called something else but a language define is calling it "Condition".

There's a number of ways to obtain/get the list of fields that are in the products table; however, I have not incorporated any of them into EP4. With as much as EP4 is able to do with any of the provided data, I have always been concerned about what information might be made tooooo easily accessible. Sure I'm certain it is overthought, but I'd rather make it as easy to do what is needed/desired by someone that has or is able to obtain the information they need than to make it easy for anyone that happens to gain access...