diff --git a/.github/.OwlBot.lock.yaml b/.github/.OwlBot.lock.yaml index a3da1b0d..a9bdb1b7 100644 --- a/.github/.OwlBot.lock.yaml +++ b/.github/.OwlBot.lock.yaml @@ -13,5 +13,5 @@ # limitations under the License. docker: image: gcr.io/cloud-devrel-public-resources/owlbot-python:latest - digest: sha256:3e3800bb100af5d7f9e810d48212b37812c1856d20ffeafb99ebe66461b61fc7 -# created: 2023-08-02T10:53:29.114535628Z + digest: sha256:fac304457974bb530cc5396abd4ab25d26a469cd3bc97cbfb18c8d4324c584eb +# created: 2023-10-02T21:31:03.517640371Z diff --git a/.gitignore b/.gitignore index b4243ced..d083ea1d 100644 --- a/.gitignore +++ b/.gitignore @@ -50,6 +50,7 @@ docs.metadata # Virtual environment env/ +venv/ # Test logs coverage.xml diff --git a/.kokoro/requirements.txt b/.kokoro/requirements.txt index 029bd342..96d593c8 100644 --- a/.kokoro/requirements.txt +++ b/.kokoro/requirements.txt @@ -113,30 +113,30 @@ commonmark==0.9.1 \ --hash=sha256:452f9dc859be7f06631ddcb328b6919c67984aca654e5fefb3914d54691aed60 \ --hash=sha256:da2f38c92590f83de410ba1a3cbceafbc74fee9def35f9251ba9a971d6d66fd9 # via rich -cryptography==41.0.3 \ - --hash=sha256:0d09fb5356f975974dbcb595ad2d178305e5050656affb7890a1583f5e02a306 \ - --hash=sha256:23c2d778cf829f7d0ae180600b17e9fceea3c2ef8b31a99e3c694cbbf3a24b84 \ - --hash=sha256:3fb248989b6363906827284cd20cca63bb1a757e0a2864d4c1682a985e3dca47 \ - --hash=sha256:41d7aa7cdfded09b3d73a47f429c298e80796c8e825ddfadc84c8a7f12df212d \ - --hash=sha256:42cb413e01a5d36da9929baa9d70ca90d90b969269e5a12d39c1e0d475010116 \ - --hash=sha256:4c2f0d35703d61002a2bbdcf15548ebb701cfdd83cdc12471d2bae80878a4207 \ - --hash=sha256:4fd871184321100fb400d759ad0cddddf284c4b696568204d281c902fc7b0d81 \ - --hash=sha256:5259cb659aa43005eb55a0e4ff2c825ca111a0da1814202c64d28a985d33b087 \ - --hash=sha256:57a51b89f954f216a81c9d057bf1a24e2f36e764a1ca9a501a6964eb4a6800dd \ - --hash=sha256:652627a055cb52a84f8c448185922241dd5217443ca194d5739b44612c5e6507 \ - --hash=sha256:67e120e9a577c64fe1f611e53b30b3e69744e5910ff3b6e97e935aeb96005858 \ - --hash=sha256:6af1c6387c531cd364b72c28daa29232162010d952ceb7e5ca8e2827526aceae \ - --hash=sha256:6d192741113ef5e30d89dcb5b956ef4e1578f304708701b8b73d38e3e1461f34 \ - --hash=sha256:7efe8041897fe7a50863e51b77789b657a133c75c3b094e51b5e4b5cec7bf906 \ - --hash=sha256:84537453d57f55a50a5b6835622ee405816999a7113267739a1b4581f83535bd \ - --hash=sha256:8f09daa483aedea50d249ef98ed500569841d6498aa9c9f4b0531b9964658922 \ - --hash=sha256:95dd7f261bb76948b52a5330ba5202b91a26fbac13ad0e9fc8a3ac04752058c7 \ - --hash=sha256:a74fbcdb2a0d46fe00504f571a2a540532f4c188e6ccf26f1f178480117b33c4 \ - --hash=sha256:a983e441a00a9d57a4d7c91b3116a37ae602907a7618b882c8013b5762e80574 \ - --hash=sha256:ab8de0d091acbf778f74286f4989cf3d1528336af1b59f3e5d2ebca8b5fe49e1 \ - --hash=sha256:aeb57c421b34af8f9fe830e1955bf493a86a7996cc1338fe41b30047d16e962c \ - --hash=sha256:ce785cf81a7bdade534297ef9e490ddff800d956625020ab2ec2780a556c313e \ - --hash=sha256:d0d651aa754ef58d75cec6edfbd21259d93810b73f6ec246436a21b7841908de +cryptography==41.0.4 \ + --hash=sha256:004b6ccc95943f6a9ad3142cfabcc769d7ee38a3f60fb0dddbfb431f818c3a67 \ + --hash=sha256:047c4603aeb4bbd8db2756e38f5b8bd7e94318c047cfe4efeb5d715e08b49311 \ + --hash=sha256:0d9409894f495d465fe6fda92cb70e8323e9648af912d5b9141d616df40a87b8 \ + --hash=sha256:23a25c09dfd0d9f28da2352503b23e086f8e78096b9fd585d1d14eca01613e13 \ + --hash=sha256:2ed09183922d66c4ec5fdaa59b4d14e105c084dd0febd27452de8f6f74704143 \ + --hash=sha256:35c00f637cd0b9d5b6c6bd11b6c3359194a8eba9c46d4e875a3660e3b400005f \ + --hash=sha256:37480760ae08065437e6573d14be973112c9e6dcaf5f11d00147ee74f37a3829 \ + --hash=sha256:3b224890962a2d7b57cf5eeb16ccaafba6083f7b811829f00476309bce2fe0fd \ + --hash=sha256:5a0f09cefded00e648a127048119f77bc2b2ec61e736660b5789e638f43cc397 \ + --hash=sha256:5b72205a360f3b6176485a333256b9bcd48700fc755fef51c8e7e67c4b63e3ac \ + --hash=sha256:7e53db173370dea832190870e975a1e09c86a879b613948f09eb49324218c14d \ + --hash=sha256:7febc3094125fc126a7f6fb1f420d0da639f3f32cb15c8ff0dc3997c4549f51a \ + --hash=sha256:80907d3faa55dc5434a16579952ac6da800935cd98d14dbd62f6f042c7f5e839 \ + --hash=sha256:86defa8d248c3fa029da68ce61fe735432b047e32179883bdb1e79ed9bb8195e \ + --hash=sha256:8ac4f9ead4bbd0bc8ab2d318f97d85147167a488be0e08814a37eb2f439d5cf6 \ + --hash=sha256:93530900d14c37a46ce3d6c9e6fd35dbe5f5601bf6b3a5c325c7bffc030344d9 \ + --hash=sha256:9eeb77214afae972a00dee47382d2591abe77bdae166bda672fb1e24702a3860 \ + --hash=sha256:b5f4dfe950ff0479f1f00eda09c18798d4f49b98f4e2006d644b3301682ebdca \ + --hash=sha256:c3391bd8e6de35f6f1140e50aaeb3e2b3d6a9012536ca23ab0d9c35ec18c8a91 \ + --hash=sha256:c880eba5175f4307129784eca96f4e70b88e57aa3f680aeba3bab0e980b0f37d \ + --hash=sha256:cecfefa17042941f94ab54f769c8ce0fe14beff2694e9ac684176a2535bf9714 \ + --hash=sha256:e40211b4923ba5a6dc9769eab704bdb3fbb58d56c5b336d30996c24fcf12aadb \ + --hash=sha256:efc8ad4e6fc4f1752ebfb58aefece8b4e3c4cae940b0994d43649bdfce8d0d4f # via # gcp-releasetool # secretstorage @@ -382,6 +382,7 @@ protobuf==3.20.3 \ # gcp-docuploader # gcp-releasetool # google-api-core + # googleapis-common-protos pyasn1==0.4.8 \ --hash=sha256:39c7e2ec30515947ff4e87fb6f456dfc6e84857d34be479c9d4a4ba4bf46aa5d \ --hash=sha256:aef77c9fb94a3ac588e87841208bdec464471d9871bd5050a287cc9a475cd0ba diff --git a/CHANGELOG.md b/CHANGELOG.md index 169ac822..69f4fb66 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,5 +1,12 @@ # Changelog +## [0.10.2-alpha](https://github.com/googleapis/python-documentai-toolbox/compare/v0.10.1-alpha...v0.10.2-alpha) (2023-10-03) + + +### Bug Fixes + +* Change `ocr_line` `<span>` to include all `ocr_word` ([#169](https://github.com/googleapis/python-documentai-toolbox/issues/169)) ([bc44dab](https://github.com/googleapis/python-documentai-toolbox/commit/bc44dabdbee713f54d3e4f9864195af924677e3a)) + ## [0.10.1-alpha](https://github.com/googleapis/python-documentai-toolbox/compare/v0.10.0-alpha...v0.10.1-alpha) (2023-09-05) diff --git a/google/cloud/documentai_toolbox/templates/hocr_document_template.xml.j2 b/google/cloud/documentai_toolbox/templates/hocr_document_template.xml.j2 index 57c7e976..63db0ada 100644 --- a/google/cloud/documentai_toolbox/templates/hocr_document_template.xml.j2 +++ b/google/cloud/documentai_toolbox/templates/hocr_document_template.xml.j2 @@ -18,9 +18,10 @@ {% set paridx = loop.index0 -%} {% for line in paragraph.lines -%} {% set lidx = loop.index0 -%} - {{ line.text }}{% for token in line.tokens -%} + {{ line.text }}{% for token in line.tokens -%} {% set tidx = loop.index0 -%} - {{ token.text }}{% endfor -%}{% endfor -%} + {{ token.text }}{% endfor -%} + {% endfor -%} {% endfor -%} {% endfor -%} diff --git a/google/cloud/documentai_toolbox/version.py b/google/cloud/documentai_toolbox/version.py index d48538a9..88a3fd38 100644 --- a/google/cloud/documentai_toolbox/version.py +++ b/google/cloud/documentai_toolbox/version.py @@ -13,4 +13,4 @@ # See the License for the specific language governing permissions and # limitations under the License. # -__version__ = "0.10.1-alpha" +__version__ = "0.10.2-alpha" diff --git a/samples/snippets/requirements-test.txt b/samples/snippets/requirements-test.txt index 1a6f9003..07a18a85 100644 --- a/samples/snippets/requirements-test.txt +++ b/samples/snippets/requirements-test.txt @@ -1,3 +1,3 @@ -pytest==7.4.1 +pytest==7.4.2 mock==5.1.0 google-cloud-bigquery==3.11.4 diff --git a/samples/snippets/requirements.txt b/samples/snippets/requirements.txt index 87a9178c..6c55b6bf 100644 --- a/samples/snippets/requirements.txt +++ b/samples/snippets/requirements.txt @@ -1,4 +1,4 @@ google-cloud-bigquery==3.11.4 -google-cloud-documentai==2.18.0 -google-cloud-storage==2.10.0 +google-cloud-documentai==2.20.0 +google-cloud-storage==2.11.0 google-cloud-documentai-toolbox==0.9.0a0 diff --git a/setup.py b/setup.py index d7772e4c..0da7c1b7 100644 --- a/setup.py +++ b/setup.py @@ -55,7 +55,7 @@ "proto-plus >= 1.22.2, <2.0.0dev; python_version>='3.11'", "grpc-google-iam-v1 >= 0.12.4, < 0.13dev", "google-cloud-bigquery >= 3.5.0, < 4.0.0dev", - "google-cloud-documentai >= 2.17.0, < 3.0.0dev", + "google-cloud-documentai >= 2.20.0, < 3.0.0dev", "google-cloud-storage >= 1.31.0, < 3.0.0dev", "google-cloud-vision >= 2.7.0, < 4.0.0dev ", "numpy >= 1.18.1", diff --git a/testing/constraints-3.7.txt b/testing/constraints-3.7.txt index 559bb547..013a679d 100644 --- a/testing/constraints-3.7.txt +++ b/testing/constraints-3.7.txt @@ -10,7 +10,7 @@ pandas==1.0.0 proto-plus==1.22.0 grpc-google-iam-v1==0.12.4 google-cloud-bigquery==3.5.0 -google-cloud-documentai==2.17.0 +google-cloud-documentai==2.20.0 google-cloud-storage==2.7.0 numpy==1.18.1 pikepdf==6.2.9 diff --git a/tests/unit/resources/toolbox_invoice_test_0_hocr.xml b/tests/unit/resources/toolbox_invoice_test_0_hocr.xml index 6885961b..0cd8e171 100644 --- a/tests/unit/resources/toolbox_invoice_test_0_hocr.xml +++ b/tests/unit/resources/toolbox_invoice_test_0_hocr.xml @@ -11,79 +11,79 @@
Invoice -Invoice -DATE: 01/01/1970 -DATE: 01/01/1970 -INVOICE: NO. 001 -INVOICE: NO. 001 -FROM: Company ABC -FROM: Company ABC -user@companyabc.com -user@companyabc.com -TO: John Doe -TO: John Doe -johndoe@email.com -johndoe@email.com -ADDRESS: 111 Main Street -ADDRESS: 111 Main Street -Anytown, USA -Anytown, USA -ADDRESS: 222 Main Street -ADDRESS: 222 Main Street -Anytown, USA -Anytown, USA -TERMS: 6 month contract -TERMS: 6 month contract -DUE: 01/01/2025 -DUE: 01/01/2025 -Item Description -Item Description -Quantity -Quantity -Price -Price -Amount -Amount -Tool A -Tool A -500 -500 -$1.00 -$1.00 -$500.00 -$500.00 -Service B -Service B -1 -1 -$900.00 -$900.00 -$900.00 -$900.00 -Resource C -Resource C -50 -50 -$12.00 -$12.00 -$600.00 -$600.00 -Subtotal -Subtotal -$2000.00 -$2000.00 -Tax -Tax -$140.00 -$140.00 -BALANCE DUE -BALANCE DUE -$2140.00 -$2140.00 -NOTES: -NOTES: -Supplies used for Project Q. -Supplies used for Project Q. -
+Invoice +DATE: 01/01/1970 +DATE: 01/01/1970 +INVOICE: NO. 001 +INVOICE: NO. 001 +FROM: Company ABC +FROM: Company ABC +user@companyabc.com +user@companyabc.com +TO: John Doe +TO: John Doe +johndoe@email.com +johndoe@email.com +ADDRESS: 111 Main Street +ADDRESS: 111 Main Street +Anytown, USA +Anytown, USA +ADDRESS: 222 Main Street +ADDRESS: 222 Main Street +Anytown, USA +Anytown, USA +TERMS: 6 month contract +TERMS: 6 month contract +DUE: 01/01/2025 +DUE: 01/01/2025 +Item Description +Item Description +Quantity +Quantity +Price +Price +Amount +Amount +Tool A +Tool A +500 +500 +$1.00 +$1.00 +$500.00 +$500.00 +Service B +Service B +1 +1 +$900.00 +$900.00 +$900.00 +$900.00 +Resource C +Resource C +50 +50 +$12.00 +$12.00 +$600.00 +$600.00 +Subtotal +Subtotal +$2000.00 +$2000.00 +Tax +Tax +$140.00 +$140.00 +BALANCE DUE +BALANCE DUE +$2140.00 +$2140.00 +NOTES: +NOTES: +Supplies used for Project Q. +Supplies used for Project Q. + \ No newline at end of file