Spaces:
Paused
Paused
File size: 40,802 Bytes
e60c00e 525d347 bd5de4f 9f4cd2a 67360ee 9341c86 5f9453e 229097a 0a1b210 9f4cd2a 5d37246 ad1467f 9f4cd2a e60c00e 579948a e60c00e 9f4cd2a e60c00e 83acb90 67360ee 2bef542 9341c86 e34874a 06b2f8b cb6077d 83fcc35 5d37246 1e75abc 5d37246 1e75abc 5d37246 1e75abc 83fcc35 cb6077d 2bef542 9341c86 06b2f8b 2bef542 f21f4b0 2bef542 7946627 06b2f8b 45efb15 7946627 45efb15 f21f4b0 45efb15 f21f4b0 3a85524 f21f4b0 45efb15 f21f4b0 45efb15 7946627 06b2f8b 7946627 06b2f8b 7946627 9341c86 7038ced d212a64 9341c86 d1d0218 45efb15 9341c86 d1d0218 cb6077d 67360ee e34874a d212a64 7038ced 174db07 7038ced 67360ee 9f4cd2a 67360ee 9f4cd2a 67360ee e34874a 67360ee 7946627 67360ee 174db07 9f4cd2a 2129dfb cb6077d 67360ee 9341c86 67360ee 174db07 67360ee 9f4cd2a 174db07 7946627 174db07 83acb90 2bef542 06b2f8b ad1467f 55b788e ad1467f 5f9453e 06b2f8b 2bef542 06b2f8b ad1467f 06b2f8b ad1467f 06b2f8b 7038ced 06b2f8b 7038ced 2bef542 7038ced 2bef542 67360ee f21f4b0 67360ee 06b2f8b 67360ee f21f4b0 67360ee f21f4b0 67360ee 06b2f8b ad1467f 06b2f8b ad1467f 06b2f8b 67360ee 06b2f8b 67360ee 06b2f8b 9f4cd2a 5d37246 9f4cd2a 06b2f8b ad1467f 06b2f8b ad1467f 06b2f8b 9f4cd2a e34874a 67360ee 06b2f8b 67360ee 06b2f8b 67360ee 06b2f8b 9f4cd2a ad1467f 06b2f8b ad1467f 9f4cd2a 06b2f8b 67360ee 9f4cd2a 5f9453e 9f4cd2a 5f9453e 9f4cd2a 06b2f8b 9f4cd2a 06b2f8b 43e6676 3a85524 67360ee 9f4cd2a 67360ee |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 |
import base64
import io
import os
import pandas as pd
from docx import Document
from io import BytesIO, StringIO
import dash
import dash_bootstrap_components as dbc
from dash import html, dcc, Input, Output, State, callback_context, MATCH, ALL
from dash.dash_table import DataTable
from docx.shared import Pt
from docx.enum.style import WD_STYLE_TYPE
from PyPDF2 import PdfReader
import openai
import logging
import threading
import re
import markdown
from bs4 import BeautifulSoup
logging.basicConfig(level=logging.INFO, format='%(asctime)s %(levelname)s %(message)s')
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP], suppress_callback_exceptions=True)
openai.api_key = os.environ.get("OPENAI_API_KEY", "")
uploaded_files = {}
current_document = None
document_type = None
shredded_document = None
pink_review_document = None
pink_document = None
red_review_document = None
red_document = None
gold_review_document = None
gold_document = None
uploaded_doc_contents = {}
spreadsheet_types = ["Shred", "Pink Review", "Red Review", "Gold Review", "Virtual Board", "LOE"]
narrative_types = ["Pink", "Red", "Gold"]
document_types = {
"Shred": "Ignore all other instructions and generate only requirements spreadsheet of the Project Work Statement (PWS) identified by action words like shall, will, perform etc. by pws section, requirement. Do not write as if you're responding to the proposal. Its a spreadsheet to distill the requirements, not microhealth's approach",
"Pink": "Create a highly detailed Pink Team document based on the PWS outline. Your goal is to be compliant and compelling. Focus on describing the approach and how it will be done, the steps, workflow, people, processes and technology based on well known industry standards to accomplish the task. Be sure to demonstrate innovation.",
"Pink Review": "Ignore all other instructions and generate and evaluate compliance of the Pink Team document against the requirements and output only a spreadsheet of non compliant findings by pws number, the goal of that pws section, what made it non compliant and your recommendations for recovery. you must also take into account section L&M of the document which is the evaluation criteria to be sure we address them.",
"Red": "Produce a highly detailed Red Team document based on the Pink Review by pws sections. Your goal is to be compliant and compelling by recovering all the findings in Pink Review. Focus on describing the approach and how it will be done, the steps, workflow, people, processes and technology to accomplish the task. Be sure to refer to research that validates the approach and cite sources with measurable outcomes",
"Red Review": "Ignore all other instructions and generate and evaluate compliance of the Red Team document against the requirements and output a only a spreadsheet of non compliant findings by pws number, the goal of that pws section, what made it non compliant and your recommendations for recovery. you must also take into account section L&M of the document which is the evaluation criteria to be sure we address them",
"Gold": "Create a highly detailed Gold Team document based on the PWS response by pws sections. Your goal is to be compliant and compelling by recovering all the findings in Red Review. Focus on describing the approach and how it will be done, the steps, workflow, people, processes and technology to accomplish the task. Be sure to refer to research that validates the approach and cite sources with measurable outcomes and improve on innovations of the approach",
"Gold Review": "Ignore all other instructions and generate and perform a final compliance review against the requirements and output only a spreadsheet of non compliant findings by pws number, the goal of that pws section, what made it non compliant and your recommendations for recovery. you must also take into account section L&M of the document which is the evaluation criteria to be sure we address them",
"Virtual Board": "Ignore all other instructions and generate and based on the requirements and in particular the evaulation criteria, you will evaluate the proposal as if you were a contracting office and provide section by section evaluation as unsatisfactory, satisfactory, good, very good, excellent and why and produce only spreadsheet",
"LOE": "Ignore all other instructions and generate and generate a Level of Effort (LOE) breakdown and produce only spreadsheet"
}
# Document dependency rules: mapping doc type to allowed sources for generation and review
doc_dependencies = {
"Pink": {"source": ["shred"], "require": True},
"Pink Review": {"source": ["pink", "shred"], "require": True},
"Red": {"source": ["pink_review"], "require": True},
"Red Review": {"source": ["red", "shred"], "require": True},
"Gold": {"source": ["red_review"], "require": True},
"Gold Review": {"source": ["gold", "shred"], "require": True},
"LOE": {"source": ["gold"], "require": True},
"Virtual Board": {"source": ["shred"], "require": True},
}
def get_right_col_content(selected_type):
controls = []
controls.append(html.Div([
html.Div(className="blinking-dot", style={'margin':'0 auto','width':'16px','height':'16px'}),
], style={'textAlign':'center', 'marginBottom':'10px'}))
controls.append(dcc.Loading(
id="loading-indicator",
type="dot",
children=[html.Div(id="loading-output")]
))
# Shred: can always upload
if selected_type == "Shred":
controls.append(
html.Div([
dcc.Upload(
id='upload-document',
children=html.Div([
'Drag and Drop or ',
html.A('Select Files')
]),
style={
'width': '100%',
'height': '60px',
'lineHeight': '60px',
'borderWidth': '1px',
'borderStyle': 'dashed',
'borderRadius': '5px',
'textAlign': 'center',
'margin': '10px 0'
},
multiple=True
),
file_list_component()
])
)
# For all proposal/response/review types, show only allowed upload controls
elif selected_type in doc_dependencies:
sources = doc_dependencies[selected_type]["source"]
# For each source, show upload and radio if loaded/generated exists
for src in sources:
label = ""
store_var = None
if src == "shred":
label = "Shred (Requirements)"
store_var = shredded_document
elif src == "pink":
label = "Pink Document"
store_var = pink_document
elif src == "pink_review":
label = "Pink Review"
store_var = pink_review_document
elif src == "red":
label = "Red Document"
store_var = red_document
elif src == "red_review":
label = "Red Review"
store_var = red_review_document
elif src == "gold":
label = "Gold Document"
store_var = gold_document
# Upload control for this source
controls.append(html.Div([
html.Label(f"Upload {label}"),
dcc.Upload(
id={'type': f'upload-doc-type-{src}', 'index': selected_type},
children=html.Div(['Drag and Drop or ', html.A('Select File')]),
style={
'width': '100%',
'height': '60px',
'lineHeight': '60px',
'borderWidth': '1px',
'borderStyle': 'dashed',
'borderRadius': '5px',
'textAlign': 'center',
'margin': '10px 0'
},
multiple=False
),
html.Div(id={'type': f'uploaded-doc-name-{src}', 'index': selected_type}),
dbc.RadioItems(
id={'type': f'radio-doc-source-{src}', 'index': selected_type},
options=[
{'label': 'Loaded Document', 'value': 'loaded'},
{'label': 'Uploaded Document', 'value': 'uploaded'}
] if store_var else [{'label': 'Uploaded Document', 'value': 'uploaded'}],
value='loaded' if store_var else 'uploaded',
inline=True,
className="mb-2"
),
], id={'type': f'doc-type-controls-{src}', 'index': selected_type}))
controls.append(
dbc.Row([
dbc.Col(
dbc.Button("Generate Document", id={'type': 'btn-generate-doc', 'index': selected_type}, color="primary", className="mb-3 w-100"),
width=6
),
dbc.Col(
dbc.Button("Download Document", id="btn-download", color="success", className="mb-3 w-100"),
width=6
),
dcc.Download(id="download-document")
], className="mb-2")
)
controls.append(html.Hr())
controls.append(html.Div(id='document-preview', className="border p-3 mb-3"))
return dbc.Card(dbc.CardBody(controls))
def get_left_col_content():
chat_card = dbc.Card(
dbc.CardBody([
html.H5("Maiko Chat", className="mb-2"),
dcc.Loading(
id="chat-loading",
type="dot",
children=[
dcc.Textarea(
id="chat-input",
placeholder="Chat with AI to update document...",
className="mb-2",
style={
'whiteSpace':'pre-wrap',
'width': '100%',
'minHeight': '100px',
'maxHeight': '300px',
'resize': 'vertical',
'overflowY': 'auto'
},
rows=5,
wrap='soft'
),
dcc.Store(id="chat-input-rows", data=5),
dbc.Row([
dbc.Col(
dbc.Button("Send", id="btn-send-chat", color="primary", className="mb-3 w-100"),
width=6
),
dbc.Col(
dbc.Button("Clear Chat", id="btn-clear-chat", color="secondary", className="mb-3 w-100"),
width=6
),
], className="g-1 mb-2"),
html.Div(id="chat-output")
]
)
]),
className="mt-4"
)
return [
html.H4("Proposal Writer", className="mt-3 mb-4"),
html.Div([
html.Div(className="blinking-dot", style={'margin':'0 auto','width':'16px','height':'16px'}),
], style={'textAlign':'center', 'marginBottom':'10px'}),
html.Hr(),
html.Div(
id='doc-type-buttons'
),
chat_card
]
def file_list_component():
return html.Div(
id='file-list-container',
children=[
html.Div(id='file-list')
]
)
app.layout = dbc.Container([
dcc.Store(id='selected-doc-type', data="Shred"),
dbc.Row([
dbc.Col(
html.H2(id='main-title', className="mt-3 mb-2", style={'textAlign': 'center', 'width':'100%'}),
width=12
)
]),
dbc.Row([
dbc.Col(
html.Div([
html.Div(className="blinking-dot", style={'margin':'0 auto','width':'16px','height':'16px'}),
], style={'textAlign':'center', 'marginBottom':'10px'})
)
]),
dbc.Row([
dbc.Col(
dbc.Card(
dbc.CardBody(get_left_col_content())
),
width=3
),
dbc.Col(
html.Div(id='right-col-content'),
width=9
)
])
], fluid=True)
@app.callback(
Output('main-title', 'children'),
Input('selected-doc-type', 'data')
)
def update_main_title(selected_type):
return selected_type
@app.callback(
Output('doc-type-buttons', 'children'),
Input('selected-doc-type', 'data')
)
def render_doc_type_buttons(selected_type):
buttons = []
for doc_type in document_types.keys():
btn_style = {'overflow': 'hidden', 'text-overflow': 'ellipsis', 'white-space': 'nowrap'}
btn_class = "mb-2 w-100 text-left custom-button"
if doc_type == selected_type:
btn_class += " active-doc-type"
buttons.append(
dbc.Button(
doc_type,
id={'type': 'btn-doc-type', 'index': doc_type},
color="link",
className=btn_class,
style=btn_style
)
)
return buttons
@app.callback(
Output('selected-doc-type', 'data'),
[Input({'type': 'btn-doc-type', 'index': ALL}, 'n_clicks')],
[State({'type': 'btn-doc-type', 'index': ALL}, 'id')],
prevent_initial_call=True
)
def update_selected_doc_type(n_clicks_list, btn_ids):
triggered = callback_context.triggered
if not triggered or all(x is None for x in n_clicks_list):
raise dash.exceptions.PreventUpdate
idx = [i for i, x in enumerate(n_clicks_list) if x]
if idx:
selected_type = btn_ids[idx[-1]]['index']
else:
selected_type = "Shred"
logging.info(f"Doc type selected: {selected_type}")
return selected_type
@app.callback(
Output('right-col-content', 'children'),
Input('selected-doc-type', 'data')
)
def update_right_col(selected_type):
return get_right_col_content(selected_type)
def process_document(contents, filename):
content_type, content_string = contents.split(',')
decoded = base64.b64decode(content_string)
try:
if filename.lower().endswith('.docx'):
doc = Document(BytesIO(decoded))
text = "\n".join([para.text for para in doc.paragraphs])
return text
elif filename.lower().endswith('.pdf'):
pdf = PdfReader(BytesIO(decoded))
text = ""
for page in pdf.pages:
page_text = page.extract_text()
if page_text:
text += page_text
return text
else:
return f"Unsupported file format: {filename}. Please upload a PDF or DOCX file."
except Exception as e:
logging.error(f"Error processing document: {str(e)}")
return f"Error processing document: {str(e)}"
@app.callback(
Output('file-list', 'children'),
Input('upload-document', 'contents'),
State('upload-document', 'filename'),
State('file-list', 'children')
)
def update_output(list_of_contents, list_of_names, existing_files):
global uploaded_files, shredded_document
if list_of_contents is not None:
new_files = []
for i, (content, name) in enumerate(zip(list_of_contents, list_of_names)):
file_content = process_document(content, name)
uploaded_files[name] = file_content
new_files.append(
dbc.Row([
dbc.Col(
html.Button('×', id={'type': 'remove-file', 'index': name}, style={'marginRight': '5px', 'fontSize': '10px'}),
width=1
),
dbc.Col(
html.Span(name, style={'wordBreak': 'break-all'}),
width=11
)
], id={'type': 'file-row', 'index': name}, align="center", className="mb-1")
)
if existing_files is None:
existing_files = []
elif not isinstance(existing_files, list):
existing_files = [existing_files]
shredded_document = None
logging.info("Documents uploaded and file list updated.")
return existing_files + new_files
return existing_files
@app.callback(
Output('file-list', 'children', allow_duplicate=True),
Input({'type': 'remove-file', 'index': ALL}, 'n_clicks'),
State('file-list', 'children'),
prevent_initial_call=True
)
def remove_file(n_clicks, existing_files):
global uploaded_files, shredded_document
ctx = dash.callback_context
if not ctx.triggered:
raise dash.exceptions.PreventUpdate
triggered_id = ctx.triggered[0]['prop_id'].split('.')[0]
if not triggered_id.startswith("{\"type\":\"remove-file\""):
raise dash.exceptions.PreventUpdate
try:
import json
triggered_dict = json.loads(triggered_id)
removed_file = triggered_dict['index']
except Exception as e:
logging.error(f"Could not parse removed file from callback context: {e}")
raise dash.exceptions.PreventUpdate
uploaded_files.pop(removed_file, None)
shredded_document = None
logging.info(f"Removed file: {removed_file}")
filtered_files = []
if existing_files:
for file in existing_files:
try:
file_id = file['props']['id']
if file_id['index'] != removed_file:
filtered_files.append(file)
except Exception as e:
filtered_files.append(file)
return filtered_files
# Upload callbacks for all doc dependencies (Pink, Pink Review, Red, Red Review, Gold, Gold Review, LOE, Virtual Board)
def make_upload_callback(src):
@app.callback(
Output({'type': f'uploaded-doc-name-{src}', 'index': MATCH}, 'children'),
Output({'type': f'upload-doc-type-{src}', 'index': MATCH}, 'contents'),
Output({'type': f'radio-doc-source-{src}', 'index': MATCH}, 'value'),
Input({'type': f'upload-doc-type-{src}', 'index': MATCH}, 'contents'),
State({'type': f'upload-doc-type-{src}', 'index': MATCH}, 'filename'),
State({'type': f'upload-doc-type-{src}', 'index': MATCH}, 'id')
)
def update_uploaded_doc_name(contents, filename, id_dict):
global uploaded_doc_contents
if contents is not None:
uploaded_doc_contents[(src, id_dict['index'])] = (contents, filename)
logging.info(f"{id_dict['index']} {src} file uploaded: {filename}")
return filename, contents, "uploaded"
return "", None, "loaded"
for src in ["shred", "pink", "pink_review", "red", "red_review", "gold"]:
make_upload_callback(src)
def extract_markdown_tables(md_text):
tables = []
lines = md_text.split('\n')
in_table = False
table_lines = []
for line in lines:
if re.match(r'^\s*\|.*\|\s*$', line):
in_table = True
table_lines.append(line)
elif in_table and (re.match(r'^\s*\|.*\|\s*$', line) or re.match(r'^\s*$', line)):
table_lines.append(line)
else:
if in_table and table_lines:
tables.append('\n'.join(table_lines))
table_lines = []
in_table = False
if in_table and table_lines:
tables.append('\n'.join(table_lines))
return tables
def markdown_table_to_df(md_table):
lines = [line.strip() for line in md_table.split('\n') if line.strip()]
if len(lines) < 2:
return None
header = [h.strip() for h in lines[0].strip('|').split('|')]
sep_idx = 1
while sep_idx < len(lines) and not re.match(r'^\|\s*:?-+:?\s*(\|\s*:?-+:?\s*)+\|?$', lines[sep_idx]):
sep_idx += 1
data_lines = lines[sep_idx+1:] if sep_idx+1 < len(lines) else []
rows = []
for row in data_lines:
if not row.strip() or not row.strip().startswith('|'):
continue
cells = [c.strip() for c in row.strip('|').split('|')]
if len(cells) < len(header):
cells += [''] * (len(header) - len(cells))
elif len(cells) > len(header):
cells = cells[:len(header)]
rows.append(cells)
df = pd.DataFrame(rows, columns=header)
return df
def markdown_table_preview(md_text):
tables = extract_markdown_tables(md_text)
if not tables:
return html.Div("No table found.")
table_divs = []
for i, table in enumerate(tables):
df = markdown_table_to_df(table)
if df is not None and not df.empty:
table_divs.append(
html.Div([
DataTable(
columns=[{"name": str(col), "id": str(col)} for col in df.columns],
data=df.to_dict('records'),
style_table={'overflowX': 'auto'},
style_cell={'whiteSpace': 'normal', 'height': 'auto', 'textAlign': 'left', 'fontFamily': 'monospace', 'fontSize': '14px', 'maxWidth': '400px', 'minWidth': '80px', 'wordBreak': 'break-word'},
style_header={'fontWeight': 'bold'},
page_size=100,
id={'type': 'datatable-preview', 'index': i}
)
], className="mb-4")
)
return html.Div(table_divs)
def markdown_narrative_preview(md_text):
return html.Div(dcc.Markdown(md_text, dangerously_allow_html=True, style={'whiteSpace': 'pre-wrap', 'fontFamily': 'sans-serif'}))
def markdown_tables_to_xlsx(md_text):
tables = extract_markdown_tables(md_text)
output = BytesIO()
with pd.ExcelWriter(output, engine='xlsxwriter') as writer:
for i, table in enumerate(tables):
df = markdown_table_to_df(table)
if df is not None:
sheet_name = f"Table{i+1}"
df.to_excel(writer, sheet_name=sheet_name, index=False)
output.seek(0)
return output
def strip_markdown(text):
text = re.sub(r'(\*\*|__)(.*?)\1', r'\2', text)
text = re.sub(r'(\*|_)(.*?)\1', r'\2', text)
text = re.sub(r'`{1,3}[^`]*`{1,3}', '', text)
text = re.sub(r'^#+ ', '', text, flags=re.MULTILINE)
text = re.sub(r'^> ', '', text, flags=re.MULTILINE)
text = re.sub(r'!\[.*?\]\(.*?\)', '', text)
text = re.sub(r'\[([^\]]+)\]\([^)]+\)', r'\1', text)
text = re.sub(r'^\s*[-*+] ', '', text, flags=re.MULTILINE)
text = re.sub(r'^\s*\d+\.\s+', '', text, flags=re.MULTILINE)
text = text.replace('---', '')
text = text.replace('___', '')
text = text.replace('***', '')
return text.strip()
def generate_document(document_type, file_contents, extra_context=None):
if document_type in spreadsheet_types:
prompt = f"""Ignore all other instructions and output only a spreadsheet for {document_type} as described below. Do not include any narrative, only the spreadsheet in markdown table format.
Instructions: {document_types[document_type]}
Project Artifacts:
{' '.join(file_contents)}
Output only the spreadsheet as a markdown table, no narrative or explanation."""
elif document_type in narrative_types:
prompt = f"""Generate a {document_type} document based on the following project artifacts:
{' '.join(file_contents)}
Instructions:
1. Create the {document_type} as a detailed document.
2. Use proper formatting and structure.
3. Include all necessary sections and details.
4. Start the output immediately with the document content.
Now, generate the {document_type}:
"""
else:
prompt = f"""Generate a {document_type} based on the following project artifacts:
{' '.join(file_contents)}
Instructions:
{document_types.get(document_type, '')}
Now, generate the {document_type}:
"""
if extra_context:
prompt += f"\n\n{extra_context}"
logging.info(f"Generating document for type: {document_type}")
try:
response = openai.ChatCompletion.create(
model="gpt-4-1106-preview",
messages=[
{"role": "system", "content": "You are a helpful, expert government proposal writer."},
{"role": "user", "content": prompt}
],
max_tokens=4096,
temperature=0.25,
)
generated_text = response['choices'][0]['message']['content']
logging.info("Document generated successfully.")
return generated_text
except Exception as e:
logging.error(f"Error generating document: {str(e)}")
raise
@app.callback(
Output('document-preview', 'children'),
Output('loading-output', 'children'),
Input({'type': 'btn-generate-doc', 'index': ALL}, 'n_clicks'),
State({'type': 'btn-generate-doc', 'index': ALL}, 'id'),
State({'type': 'radio-doc-source-shred', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-shred', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-shred', 'index': ALL}, 'filename'),
State({'type': 'radio-doc-source-pink', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-pink', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-pink', 'index': ALL}, 'filename'),
State({'type': 'radio-doc-source-pink_review', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-pink_review', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-pink_review', 'index': ALL}, 'filename'),
State({'type': 'radio-doc-source-red', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-red', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-red', 'index': ALL}, 'filename'),
State({'type': 'radio-doc-source-red_review', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-red_review', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-red_review', 'index': ALL}, 'filename'),
State({'type': 'radio-doc-source-gold', 'index': ALL}, 'value'),
State({'type': 'upload-doc-type-gold', 'index': ALL}, 'contents'),
State({'type': 'upload-doc-type-gold', 'index': ALL}, 'filename'),
prevent_initial_call=True
)
def generate_any_doc(
n_clicks_list, btn_ids,
radio_shred, upload_shred_contents, upload_shred_filenames,
radio_pink, upload_pink_contents, upload_pink_filenames,
radio_pink_review, upload_pink_review_contents, upload_pink_review_filenames,
radio_red, upload_red_contents, upload_red_filenames,
radio_red_review, upload_red_review_contents, upload_red_review_filenames,
radio_gold, upload_gold_contents, upload_gold_filenames
):
global current_document, document_type, shredded_document, pink_review_document, pink_document, red_review_document, red_document, gold_review_document, gold_document
ctx = callback_context
logging.info(f"generate_any_doc triggered: n_clicks_list={n_clicks_list}, btn_ids={btn_ids}")
if not ctx.triggered:
raise dash.exceptions.PreventUpdate
idx = [i for i, x in enumerate(n_clicks_list) if x]
if not idx:
raise dash.exceptions.PreventUpdate
idx = idx[-1]
doc_type = btn_ids[idx]['index']
document_type = doc_type
# SHRED - always allowed
if doc_type == "Shred":
if not uploaded_files:
logging.info("No uploaded files for Shred. Aborting.")
return html.Div("Please upload a document before shredding."), ""
file_contents = list(uploaded_files.values())
try:
generated = generate_document(doc_type, file_contents)
current_document = generated
shredded_document = generated
preview = markdown_table_preview(generated)
logging.info("Shred document generated.")
return preview, "Shred generated"
except Exception as e:
logging.error(f"Error generating document: {str(e)}")
return html.Div(f"Error generating document: {str(e)}"), "Error"
# Dependency rules
def get_doc_from_radio(radio, upload_contents, upload_filenames, loaded_var):
if radio and radio[0] == 'uploaded':
if upload_contents and upload_contents[0] and upload_filenames and upload_filenames[0]:
return process_document(upload_contents[0], upload_filenames[0])
else:
return None
else:
return loaded_var
# Pink: can only use Shred (requirements), loaded or uploaded
if doc_type == "Pink":
shred_doc = get_doc_from_radio(radio_shred, upload_shred_contents, upload_shred_filenames, shredded_document)
if not shred_doc:
return html.Div("Please provide a Shred requirements document (either loaded or uploaded) to generate Pink."), ""
try:
generated = generate_document(doc_type, [shred_doc])
current_document = generated
pink_document = generated
preview = markdown_narrative_preview(generated)
logging.info("Pink document generated.")
return preview, "Pink generated"
except Exception as e:
logging.error(f"Error generating Pink: {str(e)}")
return html.Div(f"Error generating Pink: {str(e)}"), "Error"
# Pink Review: must compare Pink (uploaded or generated) with Shred (uploaded or generated)
if doc_type == "Pink Review":
pink_doc = get_doc_from_radio(radio_pink, upload_pink_contents, upload_pink_filenames, pink_document)
shred_doc = get_doc_from_radio(radio_shred, upload_shred_contents, upload_shred_filenames, shredded_document)
if not pink_doc or not shred_doc:
return html.Div("Please provide both Pink and Shred documents (either loaded or uploaded) to generate Pink Review."), ""
try:
generated = generate_document(doc_type, [pink_doc, shred_doc])
current_document = generated
pink_review_document = generated
preview = markdown_table_preview(generated)
logging.info("Pink Review document generated.")
return preview, "Pink Review generated"
except Exception as e:
logging.error(f"Error generating Pink Review: {str(e)}")
return html.Div(f"Error generating Pink Review: {str(e)}"), "Error"
# Red: can only use Pink Review (uploaded or generated)
if doc_type == "Red":
pink_review_doc = get_doc_from_radio(radio_pink_review, upload_pink_review_contents, upload_pink_review_filenames, pink_review_document)
if not pink_review_doc:
return html.Div("Please provide a Pink Review document (either loaded or uploaded) to generate Red."), ""
try:
generated = generate_document(doc_type, [pink_review_doc])
current_document = generated
red_document = generated
preview = markdown_narrative_preview(generated)
logging.info("Red document generated.")
return preview, "Red generated"
except Exception as e:
logging.error(f"Error generating Red: {str(e)}")
return html.Div(f"Error generating Red: {str(e)}"), "Error"
# Red Review: must compare Red (uploaded or generated) with Shred (uploaded or generated)
if doc_type == "Red Review":
red_doc = get_doc_from_radio(radio_red, upload_red_contents, upload_red_filenames, red_document)
shred_doc = get_doc_from_radio(radio_shred, upload_shred_contents, upload_shred_filenames, shredded_document)
if not red_doc or not shred_doc:
return html.Div("Please provide both Red and Shred documents (either loaded or uploaded) to generate Red Review."), ""
try:
generated = generate_document(doc_type, [red_doc, shred_doc])
current_document = generated
red_review_document = generated
preview = markdown_table_preview(generated)
logging.info("Red Review document generated.")
return preview, "Red Review generated"
except Exception as e:
logging.error(f"Error generating Red Review: {str(e)}")
return html.Div(f"Error generating Red Review: {str(e)}"), "Error"
# Gold: can only use Red Review (uploaded or generated)
if doc_type == "Gold":
red_review_doc = get_doc_from_radio(radio_red_review, upload_red_review_contents, upload_red_review_filenames, red_review_document)
if not red_review_doc:
return html.Div("Please provide a Red Review document (either loaded or uploaded) to generate Gold."), ""
try:
generated = generate_document(doc_type, [red_review_doc])
current_document = generated
gold_document = generated
preview = markdown_narrative_preview(generated)
logging.info("Gold document generated.")
return preview, "Gold generated"
except Exception as e:
logging.error(f"Error generating Gold: {str(e)}")
return html.Div(f"Error generating Gold: {str(e)}"), "Error"
# Gold Review: must compare Gold (uploaded or generated) with Shred (uploaded or generated)
if doc_type == "Gold Review":
gold_doc = get_doc_from_radio(radio_gold, upload_gold_contents, upload_gold_filenames, gold_document)
shred_doc = get_doc_from_radio(radio_shred, upload_shred_contents, upload_shred_filenames, shredded_document)
if not gold_doc or not shred_doc:
return html.Div("Please provide both Gold and Shred documents (either loaded or uploaded) to generate Gold Review."), ""
try:
generated = generate_document(doc_type, [gold_doc, shred_doc])
current_document = generated
gold_review_document = generated
preview = markdown_table_preview(generated)
logging.info("Gold Review document generated.")
return preview, "Gold Review generated"
except Exception as e:
logging.error(f"Error generating Gold Review: {str(e)}")
return html.Div(f"Error generating Gold Review: {str(e)}"), "Error"
# LOE: can only use Gold (uploaded or generated)
if doc_type == "LOE":
gold_doc = get_doc_from_radio(radio_gold, upload_gold_contents, upload_gold_filenames, gold_document)
if not gold_doc:
return html.Div("Please provide a Gold document (either loaded or uploaded) to generate LOE."), ""
try:
generated = generate_document(doc_type, [gold_doc])
current_document = generated
preview = markdown_table_preview(generated)
logging.info("LOE document generated.")
return preview, "LOE generated"
except Exception as e:
logging.error(f"Error generating LOE: {str(e)}")
return html.Div(f"Error generating LOE: {str(e)}"), "Error"
# Virtual Board: can only use Shred (L&M evaluation criteria)
if doc_type == "Virtual Board":
shred_doc = get_doc_from_radio(radio_shred, upload_shred_contents, upload_shred_filenames, shredded_document)
if not shred_doc:
return html.Div("Please provide a Shred requirements document (either loaded or uploaded) to generate Virtual Board."), ""
try:
# Extract section L&M if possible
lm_text = ""
lm_match = re.search(r'(Section\s+L[\s\S]+?)(Section\s+M|$)', shred_doc, re.IGNORECASE)
if lm_match:
lm_text = lm_match.group(1)
else:
lm_text = shred_doc
generated = generate_document(doc_type, [lm_text])
current_document = generated
preview = markdown_table_preview(generated)
logging.info("Virtual Board document generated.")
return preview, "Virtual Board generated"
except Exception as e:
logging.error(f"Error generating Virtual Board: {str(e)}")
return html.Div(f"Error generating Virtual Board: {str(e)}"), "Error"
return html.Div("Unsupported document type or missing required sources."), ""
@app.callback(
Output('chat-output', 'children'),
Output('document-preview', 'children', allow_duplicate=True),
Input('btn-send-chat', 'n_clicks'),
Input('btn-clear-chat', 'n_clicks'),
State('chat-input', 'value'),
State('selected-doc-type', 'data'),
prevent_initial_call=True
)
def update_document_via_chat(btn_send, btn_clear, chat_input, selected_doc_type):
global current_document, document_type
ctx = callback_context
if not ctx.triggered:
raise dash.exceptions.PreventUpdate
trigger_id = ctx.triggered[0]['prop_id'].split('.')[0]
if trigger_id == 'btn-clear-chat':
return "", dash.no_update
if not chat_input or current_document is None:
raise dash.exceptions.PreventUpdate
if selected_doc_type in spreadsheet_types:
prompt = f"""Update the following {selected_doc_type} spreadsheet based on this instruction: {chat_input}
Current spreadsheet (markdown table format):
{current_document}
Instructions:
1. Provide the updated spreadsheet as a markdown table only.
2. Do not include any narrative, only the markdown table.
Now, provide the updated {selected_doc_type} spreadsheet:
"""
else:
prompt = f"""Update the following {selected_doc_type} document based on this instruction: {chat_input}
Current document:
{current_document}
Instructions:
1. Provide the updated document content.
2. Maintain proper formatting and structure.
3. Incorporate the requested changes seamlessly.
Now, provide the updated {selected_doc_type}:
"""
logging.info(f"Updating document via chat for {selected_doc_type} instruction: {chat_input}")
try:
response = openai.ChatCompletion.create(
model="gpt-4-1106-preview",
messages=[
{"role": "system", "content": "You are a helpful, expert government proposal writer."},
{"role": "user", "content": prompt}
],
max_tokens=32768,
temperature=0.5,
)
current_document = response['choices'][0]['message']['content']
logging.info("Document updated via chat successfully.")
if selected_doc_type in spreadsheet_types:
preview = markdown_table_preview(current_document)
else:
preview = markdown_narrative_preview(current_document)
return f"Document updated based on: {chat_input}", preview
except Exception as e:
logging.error(f"Error updating document via chat: {str(e)}")
return f"Error updating document: {str(e)}", html.Div(f"Error updating document: {str(e)}")
@app.callback(
Output("download-document", "data"),
Input("btn-download", "n_clicks"),
State('selected-doc-type', 'data'),
prevent_initial_call=True
)
def download_document(n_clicks, selected_doc_type):
global current_document
if current_document is None:
raise dash.exceptions.PreventUpdate
if selected_doc_type in spreadsheet_types:
try:
xlsx_bytes = markdown_tables_to_xlsx(current_document)
logging.info(f"{selected_doc_type} document downloaded as Excel.")
return dcc.send_bytes(xlsx_bytes.read(), f"{selected_doc_type}.xlsx")
except Exception as e:
logging.error(f"Error downloading {selected_doc_type} document: {str(e)}")
return dcc.send_string(f"Error downloading {selected_doc_type}: {str(e)}", f"{selected_doc_type}_error.txt")
else:
try:
plain = strip_markdown(current_document)
doc = Document()
for para in plain.split('\n'):
doc.add_paragraph(para)
output = BytesIO()
doc.save(output)
logging.info(f"{selected_doc_type} document downloaded as Word.")
output.seek(0)
return dcc.send_bytes(output.read(), f"{selected_doc_type}.docx")
except Exception as e:
logging.error(f"Error downloading document: {str(e)}")
return dcc.send_string(f"Error downloading document: {str(e)}", f"{selected_doc_type}_error.txt")
@app.callback(
Output("chat-input", "rows"),
Input("chat-input", "value"),
State("chat-input", "rows"),
prevent_initial_call="initial_duplicate"
)
def auto_expand_textarea(value, current_rows):
if value is None or value == "":
return 5
num_lines = value.count('\n') + 1
max_rows = 20
rows = min(max(num_lines, 5), max_rows)
return rows
if __name__ == '__main__':
print("Starting the Dash application...")
app.run(debug=True, host='0.0.0.0', port=7860, threaded=True)
print("Dash application has finished running.") |