Provides dockerized access to:
- Tesseract (via pytesseract)
- Spacy
NOTE: this is built for local development, and not really suited to production environments yet.
Clone this repo.
docker-compose build
docker-compose up
These endpoints are available at: http://localhost:5000
Uses tesseract to retrieve the text from an image.
{
"file": A source pdf file
}
| Name | Type | Description |
|---|---|---|
| file | file resource | The file to process for text. |
| psm | page segmentation mode | Which psm mode to use. Defaults to 3. |
| oem | engine mode | Which engine mode to use. Defaults to 3. |
Example request using GuzzleHttp:
$body = fopen('/path/to/file', 'r');
$r = $client->request('POST', 'http://localhost:5000/api/ocr', ['file' => $body]);
Example response:
{
"lines": [
"The (quick) [brown] {fox} jumps!",
"Over the $43,456.78 <lazy> #90 dog",
"& duck/goose, as 12.5% of E-mail",
"from [email protected] is spam.",
"Der ,.schnelle” braune Fuchs springt",
"iiber den faulen Hund. Le renard brun",
"«rapide» saute par-dessus le chien",
"paresseux. La volpe marrone rapida",
"salta sopra il cane pigro. El zorro",
"marron rapido salta sobre el perro",
"perezoso. A raposa marrom rapida",
"salta sobre o céo preguicoso."
],
"text": "The (quick) [brown] {fox} jumps!\nOver the $43,456.78 <lazy> #90 dog\n& duck/goose, as 12.5% of E-mail\nfrom [email protected] is spam.\nDer ,.schnelle” braune Fuchs springt\niiber den faulen Hund. Le renard brun\n«rapide» saute par-dessus le chien\nparesseux. La volpe marrone rapida\nsalta sopra il cane pigro. El zorro\nmarron rapido salta sobre el perro\nperezoso. A raposa marrom rapida\nsalta sobre o céo preguicoso."
}
Detailed example, specifying a different psm:
<?php
use GuzzleHttp\Client;
require __DIR__ . '/vendor/autoload.php';
$url = 'http://localhost:5000/api/ocr';
$client = new \GuzzleHttp\Client();
$options = [
'multipart' => [
[
'contents' => fopen('./pdf/H3485-2019-01-09-introduced.pdf', 'r'),
'filename' => 'H3485-2019-01-09-introduced.pdf',
],
[
'name' => 'psm',
'contents' => json_encode('11')
]
],
];
$response = $client->post($url, $options);
echo $response->getStatusCode(); // 200
echo"<pre>";
echo $response->getBody();
echo"</pre>";
Response:
{
"lines": [
"SOUTH CAROLINA REVENUE AND FISCAL AFFAIRS OFFICE",
"STATEMENT OF ESTIMATED FISCAL IMPACT",
"(803)734-0640 » RFA.SC.GOV/IMPACTS",
"Bill Number:",
"H. 3020",
"Introduced on January 8, 2019",
"Author:",
"MeCravy",
"Subject:",
"SC Fetal Heartbeat Protection from Abortion Act",
"Requestor:",
"House Judiciary",
"RFA Analyst(s):",
"Griffith and Gardner",
"Impact Date:",
"April 4, 2019",
"Fiscal Impact Summary",
"This bill will have no expenditure impact on the General Fund, Federal Funds, or Other Funds",
],
"text": "SOUTH CAROLINA REVENUE AND FISCAL AFFAIRS OFFICE STATEMENT OF ESTIMATED FISCAL IMPACT (803)734-0640 » RFA.SC.GOV/IMPACTS Bill Number: H. 3020 Introduced on January 8, 2019 Author: MeCravy Subject: SC Fetal Heartbeat Protection from Abortion Act Requestor: House Judiciary RFA Analyst(s): Griffith and Gardner Impact Date: April 4, 2019 Fiscal Impact Summary This bill will have no expenditure impact on the General Fund, Federal Funds, or Other Funds becat any additional expenses relating to the promulgation..."
}
Uses Spacy nlp to retrieve information about text.
Can provide an image, pdf or a text string to these endpoints. You must provide one or the other of these. If both are present, the pdf will be used as the source.
{
"file": A source pdf file, <br/>
"text": Text to process (instead of a file).
}
| Name | Type | Description |
|---|---|---|
| file | file resource | The file to process for text. |
| text | Text string | The text to process. |
Example response:
{
"sentences": [
"SOUTH CAROLINA REVENUE AND FISCAL AFFAIRS OFFICE\n\n",
"STATEMENT OF ESTIMATED FISCAL IMPACT\n",
"(803)734-0640 » RFA.SC.GOV/IMPACTS,\n\n \n\n \n\n",
"Bill Number: H. 3020 _",
"Introduced on January 8, 2019\n\n",
"Author: MeCravy\n",
"Subject:",
"Income Tax Credit, Rehabilitation Expenditures for Historic Structures",
"Requestor:",
"House Judiciary\n\nRFA Analyst(s)",
": Griffith and Gardner\nImpact Date:___April 4, 2019\n\n \n\n",
"Fiscal Impact Summary\n\n",
]
}
See /samples/sample.php for an example using GuzzleHttp.
{
"file": A source pdf file, <br/>
"text": On July 4th, 1998 a large disc shaped object was seen hovering over New York City. Spotted by David Templer, he assumed it was a UFO. However, it turned out to be a hot air balloon.
}
NOTE: provide EITHER a file or text:
| Name | Type | Description |
|---|---|---|
| file | file resource | The file to process for text. |
| text | Text string | The text to process (instead of a file). |
Example response:
{
"entities": [
[
"July 4th, 1998",
12,
26,
"DATE"
],
[
"New York City",
77,
90,
"GPE"
],
[
"David Templer",
103,
115,
"PERSON"
]
]
}
Proivdes PoS (part of speech) tagging using Spacy.
@see https://spacy.io/usage/linguistic-features#pos-tagging
@see https://spacy.io/api/annotation#pos-tagging
{
"file": A source pdf file, <br/>
"text": On July 4th, 1998 a large disc shaped object was seen hovering over New York City. Spotted by David Templer, he assumed it was a UFO. However, it turned out to be a hot air balloon.
}
NOTE: provide EITHER a file or text:
| Name | Type | Description |
|---|---|---|
| file | file resource | The file to process for text. |
| text | Text string | The text to process (instead of a file). |
Example response, with the following data for each item (for the first item below):
- token text - "Amended"
- token lemma - "ammend"
- token PoS - "VERB" (https://spacy.io/api/annotation#pos-universal)
- token tag - "VBN" (verb, non-3rd person singular present, https://spacy.io/api/annotation#pos-en)
- token dependency - "ROOT" (https://spacy.io/api/annotation#dependency-parsing-english)
{
"tokens": [
[
"Amended",
"amend",
"VERB",
"VBN",
"ROOT"
],
[
"by",
"by",
"ADP",
"IN",
"agent"
],
[
"the",
"the",
"DET",
"DT",
"det"
],
[
"House",
"House",
"PROPN",
"NNP",
"pobj"
],
[
"of",
"of",
"ADP",
"IN",
"prep"
],
[
"Representatives",
"Representatives",
"PROPN",
"NNPS",
"pobj"
],
[
"on",
"on",
"ADP",
"IN",
"prep"
],
[
"February",
"February",
"PROPN",
"NNP",
"pobj"
],
[
"28",
"28",
"NUM",
"CD",
"nummod"
],
[
",",
",",
"PUNCT",
",",
"punct"
],
[
"2020",
"2020",
"NUM",
"CD",
"nummod"
]
]
}
Proivdes PoS (part of speech) tagging and Named Entity Recognition using Spacy.
{
"file": A source pdf file, <br/>
"text": Amended by the House of Representatives on February 28, 2020.
}
NOTE: provide EITHER a file or text:
| Name | Type | Description |
|---|---|---|
| file | file resource | The file to process for text. |
| text | Text string | The text to process (instead of a file). |
Sample Output:
{
"entities": [
[
"the House of Representatives",
11,
39,
"ORG"
],
[
"February 28, 2020",
43,
60,
"DATE"
]
],
"tokens": [
[
"Amended",
"amend",
"VERB",
"VBN",
"ROOT"
],
[
"by",
"by",
"ADP",
"IN",
"agent"
],
[
"the",
"the",
"DET",
"DT",
"det"
],
[
"House",
"House",
"PROPN",
"NNP",
"pobj"
],
[
"of",
"of",
"ADP",
"IN",
"prep"
],
[
"Representatives",
"Representatives",
"PROPN",
"NNPS",
"pobj"
],
[
"on",
"on",
"ADP",
"IN",
"prep"
],
[
"February",
"February",
"PROPN",
"NNP",
"pobj"
],
[
"28",
"28",
"NUM",
"CD",
"nummod"
],
[
",",
",",
"PUNCT",
",",
"punct"
],
[
"2020",
"2020",
"NUM",
"CD",
"nummod"
]
]
}