public void LoadFromTextCopiedOfPdf(string text)
{
// Reading all codes that exist in the text, so that we can compare with
// the list of processed codes to know if any is missing at the end.
var codigosExistentes = new HashSet <string>();
var matchCodigos = Regex.Matches(text, @"\d\s?\.\s?\d\d\s?\.\s?\d\d\s?\.\s?\d\d\s?\-\s?\d\d?");
foreach (Match eachMatch in matchCodigos)
{
codigosExistentes.Add(eachMatch.Value);
}
var codigosUsados = new HashSet <string>();
// Regexes that are going to be used.
var regexCapitulo = new Regex(@"^\s*(?<NUM>\d+)?\s*CAPÍTULO\s*(?<NUM>\d+)?\s*(?<NAME>.*?)\s*$", RegexOptions.IgnoreCase | RegexOptions.Singleline);
var regexProcedimentoTipo = new Regex(
@"^([\s\w/\-]+?)\s+(\d\.\d\d\.\d\d\.\d\d\-\d)$",
RegexOptions.IgnoreCase);
var regexProcedimentoSubtipo = new Regex(
@"^(.*?)\s+\(\s?(\d\.\d\d\.\d\d\.\d\d\-\d)\)$",
RegexOptions.IgnoreCase);
Regex regexProcedimentosCapitulo = null; // this one is dynamic
var regexColumnTitles = new Regex(
@"^(?:(?:(Código|Procedimentos|Porte|Custo|Oper\.|Nº de|Aux\.?|Anest\.|Inc\.|Filme|ou Doc|UR)|(\w+\.?))(?:\s+|$))+$",
RegexOptions.IgnoreCase);
var mapColNamePattern = new Dictionary <string, string>
{
{ "código", null },
{ "procedimentos", null },
{ "porte", @"(?<PORTE>(?:(?:[\d,]+\s*de\s*)?\d+[a-z]+)|-)" },
{ "custo oper", @"(?<CUSTO>-|\d+,\d{2,3})" },
{ "porte anest", @"(?<ANEST>-|\d+)" },
{ "nº de aux", @"(?<AUX>-|\d+)" },
{ "ur", @"(?<UR>\*)" },
{ "filme ou doc", @"(?<DOC>-|\d+,\d{4})" },
{ "inc", @"(?<INC>-|\d+)" },
};
const string patternProcLine = @"
^
(?<CODE>\d\.\d\d\.\d\d\.\d\d\-\d)?\s*
(?<NAME>.+?)\s*
(?:
(?<DOTS>[\.\s]*\.\s*)
{0}
)?
\s*
(?:(?=\r\n|\r|\n|$))
";
Regex regexProcLine = null;
var regexAnexo = new Regex(@"^(OBSERVAÇÕES.*?|OBSERVAÇão.*?|INSTRUÇÕES.*?)\s*\:?$", RegexOptions.IgnoreCase);
var regexPageFooter = new Regex(
@"^(?<PAGE>\d+)?\s*Classificação Brasileira Hierarquizada de Procedimentos Médicos - (?:\d{4}|(?:\d{1,2}.*?edi[cç][aã]o))\s*(?<PAGE>\d+)?$",
RegexOptions.IgnoreCase);
var regexNumber = new Regex(@"^\d+$");
// State variables.
Pagina curPagina = null;
Capitulo curCapitulo = null;
ProcTipo curProcTipo = null;
ProcSubtipo curProcSubtipo = null;
TextoAnexo curAnexo = null; // this can span multiples pages
var builder = new StringBuilder(10000);
int pageNum = 0;
string curGroupName = null; // this happens at page 154 e 155
bool isExplicitAttachment = false; // this happens at page 202
// Processing all pages and lines of the text.
var splitPages = text.Split('\f');
for (int itPage = 0; itPage < splitPages.Length; itPage++)
{
var pageText = splitPages[itPage];
var matchExplicitMissing = Regex.Matches(pageText, @"####EXPLICIT-MISSING:(.*?)(?=####|\r\n|\n|\r|$)");
foreach (var eachMatch in matchExplicitMissing.Cast <Match>())
{
codigosUsados.Add(eachMatch.Groups[1].Value);
}
pageText = Regex.Replace(pageText, @" *####(?:ERRO|EXPLICIT-MISSING).*?(?=\r|\n|$)", "");
curPagina = this.CreatePagina(itPage, Regex.Replace(pageText, @" *####.*?(?=\r|\n|$)", ""));
var matchCapitulo = regexCapitulo.Match(pageText);
if (matchCapitulo.Success)
{
// If there is an attachment being read, we need to finalize it.
if (curAnexo != null)
{
curAnexo.Texto = builder.ToString().Trim();
CheckForMultipleLines(curAnexo.Texto);
builder.Clear();
curAnexo = null;
}
// If this page is the chapter title we must create a new chapter object.
int num = int.Parse(matchCapitulo.Groups["NUM"].Value);
string name = Regex.Replace(matchCapitulo.Groups["NAME"].Value, @"\s+", " ");
curCapitulo = this.CreateCapitulo(num, name, curPagina);
// Setting the new value of the regex, that depends on the name of the chapter.
regexProcedimentosCapitulo = new Regex(
string.Format(@"^{0}$", Regex.Replace(name, @"\s+", @"\s+")),
RegexOptions.IgnoreCase);
continue;
}
if (curCapitulo != null)
{
// State variables to read the lines.
bool canReadCapitulo = true;
bool canReadProcTipo = true;
bool canReadColumnTitle = true;
int columnTitleLines = 0; // lines of text devoted to the column titles.
bool foundPageFooter = false;
bool foundPageNum = false;
var splitLines = Regex.Split(pageText, @"\r\n|\r|\n");
for (int itLine = 0; itLine < splitLines.Length; itLine++)
{
var lineText = splitLines[itLine].Trim();
if (lineText.Contains("####ANEXO-INICIO"))
{
isExplicitAttachment = true;
}
if (lineText.Contains("####ANEXO-FIM"))
{
isExplicitAttachment = false;
lineText = Regex.Replace(lineText, @" *####.*?(?=\r|\n|$)", "");
}
if (string.IsNullOrEmpty(lineText))
{
// We use blank lines only in the attachments, such as observations and instructions.
bool canReadAnexo = !canReadCapitulo && !canReadProcTipo && !canReadColumnTitle && !foundPageFooter;
if (curAnexo != null && canReadAnexo)
{
builder.AppendLine();
}
continue;
}
if (lineText.Contains("####GROUP"))
{
lineText = Regex.Replace(lineText, @" *####.*?(?=\r|\n|$)", "");
curGroupName = lineText;
continue;
}
if (lineText.Contains("####ANEXO-CAPITULO"))
{
// This item has no specific code.
lineText = Regex.Replace(lineText, @" *####.*?(?=\r|\n|$)", "");
curAnexo = this.CreateAnexo(
null,
lineText.Trim(':').Trim().ToUpperInvariant(),
curPagina,
null,
null);
curAnexo.Capitulo = curCapitulo;
curCapitulo.Anexos.Add(curAnexo);
builder.Clear();
builder.AppendLine(lineText);
continue;
}
// Chapter heading that exists in every page.
// When we find this heading, then we set the page's chapter property.
if (canReadCapitulo)
{
var matchProcCapitulo = regexProcedimentosCapitulo.Match(lineText);
if (matchProcCapitulo.Success)
{
// Seting the chapter that the page belongs to.
curPagina.Capitulo = curCapitulo;
canReadCapitulo = false;
continue;
}
}
// Reading the ProcTipo heading that exists after the Chapter heading.
// This heading contains the level-2 node of the CBHPM tree.
if (canReadProcTipo)
{
var matchProcedimentoTipo = regexProcedimentoTipo.Match(lineText);
if (matchProcedimentoTipo.Success)
{
// All pages inside the same ProcTipo have the same heading,
// so we use a method that creates the object if it does not exist,
// or returns the existing one.
var curProcTipo2 = this.GetOrCreateProcTipo(
matchProcedimentoTipo.Groups[2].Value,
matchProcedimentoTipo.Groups[1].Value,
curPagina,
curCapitulo);
if (curProcTipo != curProcTipo2)
{
// If there is an attachment being read, we need to finalize it.
if (curAnexo != null)
{
curAnexo.Texto = builder.ToString().Trim();
CheckForMultipleLines(curAnexo.Texto);
builder.Clear();
curAnexo = null;
}
}
curProcTipo = curProcTipo2;
codigosUsados.Add(curProcTipo.Codigo);
canReadProcTipo = false;
continue;
}
}
// All pages that contains medical procedures, have the titles of the columns.
// We need these columns to know what is inside each column, and set the values
// correctly.
if (canReadColumnTitle)
{
var matchColumnTitles = regexColumnTitles.Match(lineText);
if (matchColumnTitles.Success && !matchColumnTitles.Groups[2].Success)
{
var columnCaptures = matchColumnTitles.Groups[1].Captures.Cast <Capture>().ToArray();
foreach (var eachCapture in columnCaptures)
{
curPagina.ColunaAdd(eachCapture.Value.TrimEnd('.'));
}
columnTitleLines++;
if (columnTitleLines > 2)
{
throw new Exception("More than 2 lines for column titles.");
}
continue;
}
if (columnTitleLines == 0)
{
throw new Exception("No column titles have been found.");
}
// If there is no more columns to read, then we create the regex to read
// each medical procedure.
var valuePatternsOfPage = new List <string>();
foreach (var eachColName in curPagina.Colunas)
{
var colPattern = mapColNamePattern[eachColName];
if (colPattern != null)
{
valuePatternsOfPage.Add(colPattern);
}
}
var patternValues = string.Format(@"(?:{0})", string.Join(@"\s+", valuePatternsOfPage));
regexProcLine = new Regex(
string.Format(patternProcLine, patternValues),
RegexOptions.IgnoreCase | RegexOptions.IgnorePatternWhitespace);
canReadColumnTitle = false;
}
// Reading the footer of the page.
// The footer can not come before the end of the page.
{
var matchPageFooter = regexPageFooter.Match(lineText);
if (matchPageFooter.Success)
{
if (foundPageFooter)
{
throw new Exception("Page footer duplicated.");
}
var strPageNum = matchPageFooter.Groups["PAGE"].Value;
if (int.TryParse(strPageNum, out pageNum))
{
if (curPagina.Numero != null && curPagina.Numero != pageNum)
{
throw new Exception("Page already has a number.");
}
if (foundPageNum)
{
throw new Exception("Page number duplicated.");
}
foundPageNum = true;
curPagina.Numero = pageNum;
pageNum = 0;
}
foundPageFooter = true;
continue;
}
var matchNumber = regexNumber.Match(lineText);
if (matchNumber.Success)
{
pageNum = int.Parse(lineText);
if (!foundPageFooter || curPagina.Numero != null && curPagina.Numero != pageNum)
{
throw new Exception("Found a lost number.");
}
if (foundPageNum)
{
throw new Exception("Page number duplicated.");
}
foundPageNum = true;
curPagina.Numero = pageNum;
pageNum = 0;
continue;
}
}
if (foundPageFooter)
{
throw new Exception("Cannot read after page footer.");
}
// Reading the remaining of the attachment, if there is a current attachment.
// The text of the attachment always start by multiple spaces.
if (curAnexo != null)
{
if (splitLines[itLine].StartsWith(" ") || isExplicitAttachment)
{
lineText = Regex.Replace(splitLines[itLine], @" *####.*?(?=\r|\n|$)", "");
builder.AppendLine(lineText);
continue;
}
}
// If there is an attachment being read, we need to finalize it.
if (curAnexo != null)
{
curAnexo.Texto = builder.ToString().Trim();
CheckForMultipleLines(curAnexo.Texto);
builder.Clear();
curAnexo = null;
}
// Reading ProcSubtipo. A page may contain multiple subtipos.
// Each subtipo exists only once in the document, it is not repeated in all
// pages like ProcTipo.
{
var subtipoText = lineText;
Match matchProcedimentoSubtipo = null;
for (int itNext = 0; itNext < 3; itNext++)
{
matchProcedimentoSubtipo = regexProcedimentoSubtipo.Match(subtipoText);
if (matchProcedimentoSubtipo.Success)
{
itLine += itNext;
break;
}
var nextLine = splitLines[itLine + itNext + 1];
if (!Regex.IsMatch(nextLine, @"^\s?\w+") || Regex.IsMatch(nextLine, @"(?<!\()\d\.\d\d\.\d\d\.\d\d-\d(?!\))"))
{
break;
}
// If not match, then we try to append the next line, and do the match again.
// Some SubtipoTexts are split in 2/3 lines.
subtipoText = string.Format("{0} {1}", subtipoText, nextLine);
}
if (matchProcedimentoSubtipo.Success)
{
curProcSubtipo = this.CreateProcSubtipo(
matchProcedimentoSubtipo.Groups[2].Value,
matchProcedimentoSubtipo.Groups[1].Value,
curPagina,
curProcTipo);
curGroupName = null;
codigosUsados.Add(curProcSubtipo.Codigo);
continue;
}
}
// Reading the medical procedures in the current page.
// Note that some of these items may not be procedures but observations.
{
var matchProcedimento = regexProcLine.Match(lineText);
if (matchProcedimento.Success && matchProcedimento.Groups["CODE"].Success)
{
string code = matchProcedimento.Groups["CODE"].Value;
var nameLine1 = matchProcedimento.Groups["NAME"].Value;
if (regexAnexo.IsMatch(nameLine1))
{
curAnexo = this.CreateAnexo(
code,
nameLine1.Trim(':').Trim().ToUpperInvariant(),
curPagina,
curProcTipo,
curProcSubtipo);
codigosUsados.Add(code);
builder.Clear();
builder.AppendLine(nameLine1);
}
else
{
builder.Clear();
builder.Append(nameLine1);
// Reading next lines until we find the end of the medical procedure.
for (; itLine < splitLines.Length;)
{
if (matchProcedimento.Groups["PORTE"].Success)
{
string procName = builder.ToString();
builder.Clear();
string procPorte = matchProcedimento.Groups["PORTE"].Value;
string procCusto = matchProcedimento.Groups["CUSTO"].Value;
string procAnest = matchProcedimento.Groups["ANEST"].Value;
string procAux = matchProcedimento.Groups["AUX"].Value;
string procUR = matchProcedimento.Groups["UR"].Value;
string procInc = matchProcedimento.Groups["INC"].Value;
string procDoc = matchProcedimento.Groups["DOC"].Value;
var proc = new Proc();
proc.Cbhpm = this;
proc.Codigo = code;
proc.Nome = procName;
proc.Porte = procPorte;
proc.CustoOper = procCusto;
proc.PorteAnest = procAnest;
proc.NumAux = procAux;
proc.Ur = procUR;
proc.Inc = procInc;
proc.FilmeOuDoc = procDoc;
proc.GrupoNoSubtipo = curGroupName;
proc.Subtipo = curProcSubtipo;
curProcSubtipo.Procedimentos.Add(proc);
proc.PaginaDeclarada = curPagina;
var simpleCode = GetCodeSimple(code);
this.Items.Add(StructTuple.Create(simpleCode, 0), proc);
codigosUsados.Add(code);
break;
}
itLine++;
lineText = splitLines[itLine].Trim();
matchProcedimento = regexProcLine.Match(lineText);
if (Regex.IsMatch(lineText, @"\d\.\d\d\.\d\d\.\d\d-\d"))
{
throw new Exception("Code must be found only in first line of medical procedure.");
}
if (!splitLines[itLine].StartsWith(" "))
{
throw new Exception("All lines after the first one of medical procedure, must start with spaces.");
}
builder.Append(' ' + matchProcedimento.Groups["NAME"].Value);
}
if (itLine >= splitLines.Length)
{
throw new Exception("Medical procedure not terminated.");
}
}
continue;
}
throw new Exception("Unknown line found.");
}
}
}
// Checking the columns of the page.
FinalizarPagina(curPagina);
// Finishing the page and checking page properties.
if (curPagina.Capitulo != curCapitulo && !string.IsNullOrEmpty(pageText))
{
throw new Exception();
}
if (curPagina.ProcTipo != curProcTipo && !string.IsNullOrEmpty(pageText))
{
throw new Exception();
}
}
// Checking codes that were not processed.
var codigosNaoUsados = new HashSet <string>(codigosExistentes);
codigosNaoUsados.ExceptWith(codigosUsados);
if (codigosNaoUsados.Any())
{
throw new Exception("There are unused codes!");
}
}