ai invoice importing with batch queries

1 files changed, 72 insertions, 6 deletions

diff --git a/src/ai_providers/openAI.cpp b/src/ai_providers/openAI.cpp
index fa2cc05..64e7a6a 100644
--- a/src/ai_providers/openAI.cpp
+++ b/src/ai_providers/openAI.cpp
@@ -14,6 +14,8 @@
 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 */
 
+#include <threads.h>
+
 #define CPPHTTPLIB_OPENSSL_SUPPORT
 #include "httplib.h"
 
@@ -22,9 +24,77 @@
 #include "logger.hpp"
 #include "importer.hpp"
 
+static bool _openAI_batch_query_with_file(char** queries, size_t query_count, char* file_id, invoice* buffer, importer::batch_query_response_handler response_handler)
+{
+    const char *api_key = administration::get_ai_service().api_key_public;
+    httplib::SSLClient cli("api.openai.com", 443);
+
+    thrd_t threads[20];
+
+    for (u32 i = 0; i < query_count; i++)
+    {       
+        auto* func = new auto([&api_key, &cli, i, &file_id, &response_handler, &buffer, &queries]() {
+            char* query_escaped = strops::prep_str_for_json(queries[i], 1000);
+
+            size_t body_size = 1000; // Ballpark
+            char* body = (char*)memops::alloc(body_size);
+            strops::format(body, body_size, 
+                "{ \"model\":\"%s\", "
+                "  \"input\": [ "
+                "  { \"role\": \"user\", "
+                "    \"content\": [ "
+                "    { \"type\": \"input_file\", \"file_id\": \"%s\" }, "
+                "    { \"type\": \"input_text\", \"text\": \"%s\" } "
+                "  ] "
+                "  }"
+                "], "
+                " \"text\":  { \"format\": { \"type\": \"json_object\" } } "
+                "}", administration::get_ai_service().model_name, file_id, query_escaped);
+
+            httplib::Headers headers;
+            headers.insert(std::make_pair("Authorization", std::string("Bearer ") + api_key));
+
+            httplib::Result res = cli.Post("/v1/responses", headers, body, "application/json");
+            memops::unalloc(body);
+
+            if (!res || res->status != 200) {
+                logger::error("ERROR Failed to query API.");
+                logger::error(res->body.c_str());
+                return 0;
+            }
+
+            char* response_body = (char*)res->body.c_str();
+            char* response = (char*)memops::alloc(5000);
+            memops::zero(response, 5000);
+            strops::copy(response, response_body, 5000);
+            
+            strops::get_json_value(response, "text", response, 5000);
+            strops::unprep_str_from_json(response);
+
+            response_handler(buffer, response);
+            
+            memops::unalloc(response);
+            memops::unalloc(query_escaped);
+            return 1;
+        });
+
+        auto trampoline = [](void* arg) -> int {
+            auto* f = static_cast<decltype(func)>(arg);
+            (*f)();
+            delete f;
+            return 0;
+        };
+        
+        thrd_create(&threads[i], trampoline, func);
+    }
+
+    for (u32 i = 0; i < query_count; i++) thrd_join(threads[i], nullptr);
+
+    return 1;
+}
+
 static bool _openAI_query_with_file(char* query, size_t query_length, char* file_id, char** response)
 {
-    #if 1
     const char *api_key = administration::get_ai_service().api_key_public;
     
     httplib::SSLClient cli("api.openai.com", 443);
@@ -58,11 +128,6 @@ static bool _openAI_query_with_file(char* query, size_t query_length, char* file
     
     strops::get_json_value(*response, "text", *response, 100000);
     *response = strops::unprep_str_from_json(*response);
-    #else
-    *response = (char*)memops::alloc(100000);
-    memops::zero(*response, 100000);
-    strops::copy(*response, "<Invoice xmlns=\"urn:oasis:names:specification:ubl:schema:xsd:Invoice-2\" xmlns:cac=\"urn:oasis:names:specification:ubl:schema:xsd:CommonAggregateComponents-2\" xmlns:cbc=\"urn:oasis:names:specification:ubl:schema:xsd:CommonBasicComponents-2\">  <cbc:CustomizationID>urn:cen.eu:en16931:2017#compliant#urn:fdc:peppol.eu:2017:poacc:billing:3.0</cbc:CustomizationID>  <cbc:ProfileID>urn:fdc:peppol.eu:2017:poacc:billing:01:1.0</cbc:ProfileID>  <cbc:ID>492043632</cbc:ID>  <cbc:IssueDate>2024-09-01</cbc:IssueDate>  <cbc:DueDate>2024-09-01</cbc:DueDate>  <cbc:InvoiceTypeCode>380</cbc:InvoiceTypeCode>  <cbc:DocumentCurrencyCode>USD</cbc:DocumentCurrencyCode>  <cac:DespatchDocumentReference>    <cbc:ID>Final invoice</cbc:ID>  </cac:DespatchDocumentReference>  <cac:AdditionalDocumentReference>    <cbc:ID></cbc:ID>    <cbc:DocumentDescription></cbc:DocumentDescription>  </cac:AdditionalDocumentReference>  <cac:OrderReference>    <cbc:ID></cbc:ID>  </cac:OrderReference>  <cac:ProjectReference>    <cbc:ID>do:team:67840ecb-44e2-472e-bc45-801bd4e1f1fe</cbc:ID>  </cac:ProjectReference>  <cbc:AccountingCost></cbc:AccountingCost>  <cac:AccountingSupplierParty>    <cac:Party>      <cbc:EndpointID schemeID=""></cbc:EndpointID>      <cac:PartyIdentification>        <cbc:ID schemeID=\"ZZZ\"></cbc:ID>      </cac:PartyIdentification>      <cac:PartyName>        <cbc:Name>DigitalOcean LLC</cbc:Name>      </cac:PartyName>      <cac:PostalAddress>        <cbc:StreetName>101 Avenue of the Americas</cbc:StreetName>        <cbc:AdditionalStreetName>2nd Floor</cbc:AdditionalStreetName>        <cbc:CityName>New York</cbc:CityName>        <cbc:PostalZone>10013</cbc:PostalZone>        <cbc:CountrySubentity>NY</cbc:CountrySubentity>        <cac:Country>          <cbc:IdentificationCode>US</cbc:IdentificationCode>        </cac:Country>      </cac:PostalAddress>      <cac:PartyTaxScheme>        <cbc:CompanyID>EU528002224</cbc:CompanyID>        <cac:TaxScheme>          <cbc:ID>VAT</cbc:ID>        </cac:TaxScheme>      </cac:PartyTaxScheme>      <cac:PartyLegalEntity>        <cbc:RegistrationName>DigitalOcean LLC</cbc:RegistrationName>      </cac:PartyLegalEntity>      <cac:Contact>        <cbc:Name></cbc:Name>        <cbc:Telephone></cbc:Telephone>        <cbc:ElectronicMail></cbc:ElectronicMail>      </cac:Contact>    </cac:Party>  </cac:AccountingSupplierParty>  <cac:AccountingCustomerParty>    <cac:Party>      <cbc:EndpointID schemeID=""></cbc:EndpointID>      <cac:PartyIdentification>        <cbc:ID schemeID=\"ZZZ\"></cbc:ID>      </cac:PartyIdentification>      <cac:PartyName>        <cbc:Name>My Team</cbc:Name>      </cac:PartyName>      <cac:PostalAddress>        <cbc:StreetName>Keerderstraat 81</cbc:StreetName>        <cbc:AdditionalStreetName></cbc:AdditionalStreetName>        <cbc:CityName>Maastricht</cbc:CityName>        <cbc:PostalZone>6226 XW</cbc:PostalZone>        <cbc:CountrySubentity>LI</cbc:CountrySubentity>        <cac:Country>          <cbc:IdentificationCode>NL</cbc:IdentificationCode>        </cac:Country>      </cac:PostalAddress>      <cac:PartyTaxScheme>        <cbc:CompanyID></cbc:CompanyID>        <cac:TaxScheme>          <cbc:ID>VAT</cbc:ID>        </cac:TaxScheme>      </cac:PartyTaxScheme>      <cac:PartyLegalEntity>        <cbc:RegistrationName></cbc:RegistrationName>      </cac:PartyLegalEntity>      <cac:Contact>        <cbc:Name></cbc:Name>        <cbc:Telephone></cbc:Telephone>        <cbc:ElectronicMail>aldrikboy@gmail.com</cbc:ElectronicMail>      </cac:Contact>    </cac:Party>  </cac:AccountingCustomerParty>  <cac:Delivery>    <cbc:ActualDeliveryDate></cbc:ActualDeliveryDate>    <cac:DeliveryLocation>      <cac:Address>        <cbc:StreetName></cbc:StreetName>        <cbc:AdditionalStreetName></cbc:AdditionalStreetName>        <cbc:CityName></cbc:CityName>        <cbc:PostalZone></cbc:PostalZone>        <cbc:CountrySubentity></cbc:CountrySubentity>        <cac:Country>          <cbc:IdentificationCode></cbc:IdentificationCode>        </cac:Country>      </cac:Address>    </cac:DeliveryLocation>    <cac:DeliveryParty>      <cac:PartyName>        <cbc:Name></cbc:Name>      </cac:PartyName>    </cac:DeliveryParty>  </cac:Delivery>  <cac:PaymentMeans>    <cbc:PaymentMeansCode></cbc:PaymentMeansCode>    <cbc:PaymentID>492043632</cbc:PaymentID>    <cac:PayeeFinancialAccount>      <cbc:ID></cbc:ID>      <cbc:Name></cbc:Name>      <cac:FinancialInstitutionBranch>        <cac:FinancialInstitution>          <cbc:ID></cbc:ID>        </cac:FinancialInstitution>      </cac:FinancialInstitutionBranch>    </cac:PayeeFinancialAccount>    <cac:PayerFinancialAccount>      <cbc:ID></cbc:ID>    </cac:PayerFinancialAccount>  </cac:PaymentMeans>  <cac:TaxTotal>    <cbc:TaxAmount currencyID=\"USD\">3.49</cbc:TaxAmount>    <cac:TaxSubtotal>      <cbc:TaxableAmount currencyID=\"USD\">15.60</cbc:TaxableAmount>      <cbc:TaxAmount currencyID=\"USD\">3.28</cbc:TaxAmount>      <cac:TaxCategory>        <cac:TaxScheme>          <cbc:ID>VAT</cbc:ID>        </cac:TaxScheme>      </cac:TaxCategory>    </cac:TaxSubtotal>    <cac:TaxSubtotal>      <cbc:TaxableAmount currencyID=\"USD\">1.00</cbc:TaxableAmount>      <cbc:TaxAmount currencyID=\"USD\">0.21</cbc:TaxAmount>      <cac:TaxCategory>        <cac:TaxScheme>          <cbc:ID>VAT</cbc:ID>        </cac:TaxScheme>      </cac:TaxCategory>    </cac:TaxSubtotal>  </cac:TaxTotal>  <cac:LegalMonetaryTotal>    <cbc:LineExtensionAmount currencyID=\"USD\">16.60</cbc:LineExtensionAmount>    <cbc:TaxExclusiveAmount currencyID=\"USD\">16.60</cbc:TaxExclusiveAmount>    <cbc:TaxInclusiveAmount currencyID=\"USD\">20.09</cbc:TaxInclusiveAmount>    <cbc:PayableAmount currencyID=\"USD\">20.09</cbc:PayableAmount>  </cac:LegalMonetaryTotal>  <cac:InvoiceLine>    <cbc:ID>1</cbc:ID>    <cbc:InvoicedQuantity unitCode=""></cbc:InvoicedQuantity>    <cbc:LineExtensionAmount currencyID=\"USD\">16.60</cbc:LineExtensionAmount>    <cac:AllowanceCharge>      <cbc:ChargeIndicator>false</cbc:ChargeIndicator>      <cbc:AllowanceChargeReason>Discount</cbc:AllowanceChargeReason>      <cbc:MultiplierFactorNumeric></cbc:MultiplierFactorNumeric>      <cbc:Amount currencyID=\"USD\"></cbc:Amount>      <cbc:BaseAmount currencyID=\"USD\"></cbc:BaseAmount>    </cac:AllowanceCharge>    <cac:Item>      <cbc:Name>Product Usage Charges</cbc:Name>      <cac:AdditionalItemProperty>        <cbc:Name>Internal Tax Rate ID</cbc:Name>        <cbc:Value></cbc:Value>      </cac:AdditionalItemProperty>      <cac:ClassifiedTaxCategory>        <cbc:ID></cbc:ID>        <cbc:Percent></cbc:Percent>        <cac:TaxScheme>          <cbc:ID>VAT</cbc:ID>        </cac:TaxScheme>      </cac:ClassifiedTaxCategory>    </cac:Item>    <cac:Price>      <cbc:PriceAmount currencyID=\"USD\"></cbc:PriceAmount>    </cac:Price>  </cac:InvoiceLine></Invoice>", 100000);
-    #endif
 
     return 1;
 }
@@ -205,5 +270,6 @@ importer::ai_provider_impl _chatgpt_api_provider = {
     "gpt-5-nano",
     _openAI_upload_file,
     _openAI_query_with_file,
+    _openAI_batch_query_with_file,
     _openAI_get_available_models,
 };
\ No newline at end of file